← Home
The Knowledge Project · April 22, 2026 · 1h 12m

Greg Brockman: Inside the 72 Hours That Almost Killed OpenAI

Greg Brockman, Ilya Sutskever, Sam Altman
Show Notes

Greg Brockman, co-founder and President of OpenAI, recounts the internal crisis that nearly destroyed the company when Sam Altman was fired by the board. He details the chaotic 72 hours that followed—from his immediate resignation to the secret "Phoenix" backup company designed at Sam's house, the pivotal moment when Ilya Sutskever's tweet shifted the power dynamic, and how the company emerged fundamentally changed. The conversation extends into OpenAI's future: whether we're in a true global AI race, how AI now writes much of OpenAI's own code, and what compute constraints mean for AGI access.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Highlights

•
AI Now Writes Most of OpenAI's Code33:24
A significant and growing percentage of OpenAI's own code is now written by AI; Brockman admits it's hard to measure exactly how much is human-written versus AI-generated.
ClaudeChatGPTGemini

Editorial

•
The Napa Offsite Technical Plan00:02:40
OpenAI's original technical strategy, established at a Napa offsite early in the company's history, has guided the company's research and product development for a full decade.
ClaudeChatGPTGemini
•
The Nonprofit-to-For-Profit Hybrid Was Forced, Not Ideological00:04:54
OpenAI's shift from pure nonprofit to a hybrid for-profit structure wasn't a matter of ideology or greed—it was an existential requirement to raise the capital needed to compete.
ClaudeChatGPTGemini
•
The 72-Hour Coup and the Phoenix Plan15:44
After Sam Altman was fired by the OpenAI board, Greg Brockman quit the same day; the next morning, he and Sam designed a "Phoenix" backup company at Sam's house—a legal structure ready to absorb OpenAI's entire team if the board didn't capitulate.
ClaudeChatGPTGemini
•
Ilya's Tweet as Institutional Leverage23:43
Chief Scientist Ilya Sutskever's public announcement that he was leaving OpenAI—made via tweet—was the moment that shifted the entire power dynamic and forced the board to back down.
ClaudeChatGPTGemini

Misc

✧Brockman was at a dinner when he got the board call firing Sam Altman—quit the same day
✧The 'Phoenix' backup company was literally designed at Sam's house the morning after the firing
✧Ilya Sutskever's tweet essentially ended the board's leverage and forced a resolution
✧OpenAI's own engineers now write code using AI—'it's hard to know what percent is not' AI-generated
✧The original Napa offsite technical plan has guided OpenAI's strategy for a full decade
✧The nonprofit-to-for-profit hybrid structure wasn't ideological—it was existential necessity
Was this useful?