Power Shift at OpenAI: Brockman Consolidates Control

Greg Brockman, OpenAI co-founder and president, recounts the 72-hour crisis after Sam Altman’s firing and the near-collapse of OpenAI, revealing why he resigned from the board and how a contingency “Phoenix” company plan was sketched at Altman’s house. Brockman details the origins of OpenAI’s three-step technical roadmap from the Napa offsite, the shift away from a pure nonprofit model, and how Ilya Sutskever’s public reaction changed events. He also discusses strategic choices such as hiding model reasoning traces, the role of AI in writing OpenAI’s own code, global AI race dynamics, compute constraints shaping AGI access, and implications for jobs and the industry.

Greg Brockman, OpenAI co-founder and president, recounts the internal crisis that nearly destroyed OpenAI during the 72 hours after Sam Altman’s firing, explaining why he resigned and describing the rapid creation of a contingency “Phoenix” company. Brockman reviews the formative Napa offsite that set OpenAI’s decade-long technical plan, why the organization shifted away from a pure nonprofit model, and how Ilya Sutskever’s public reaction altered events. He also discusses strategic issues: whether AI development is in a global race, the growing share of OpenAI code written with AI assistance, the decision to stop exposing model reasoning traces, compute constraints shaping AGI access, and implications for jobs and society. This episode sheds light on governance, product choices, and competitive pressure at a leading AI firm.