OpenAI #16: A History and a Proposal
Short summary
This article breaks down The New Yorker’s 18,000-word exposé on Sam Altman and OpenAI, detailing boardroom coups, safety disputes, secret pacts, and clashes with Musk, Amodei, and others. It then covers OpenAI’s policy “new deal” proposal and their acquisition of TBPN.
- Sutskever compiled seventy pages of Slack messages before Altman's firing, arguing he and Brockman shouldn't lead the company
- Musk, Altman, and Brockman had a secret pact that Altman would step down if both Brockman and Sutskever asked—Musk allegedly broke it by building a shadow leadership team
- A merger-blocking clause was quietly inserted into OpenAI's charter during Microsoft's investment, contradicting a "merge-and-assist" safety provision Amodei had demanded—Altman denied its existence until forced to read it aloud
- Summers and Taylor's promised investigation into Altman was narrowed to only assess criminality, produced no public report, and cleared him without most board members ever seeing a briefing
Full summary
OpenAI’s board drama unfolds as a series of power plays and conflicting memories. When Sam Altman was abruptly fired, board member Ilya Sutskever had already collected seventy pages of disappearing Slack messages to argue that Altman and Greg Brockman shouldn’t lead the company. Altman retreated to his $27 million San Francisco home, set up a “government-in-exile” war room, then rallied support with 12-hour calls. He later told the board “you need to chill” over Negronis at 6 p.m., but pressed on until he was reinstated. The New Yorker piece by Ronan Farrow and Andrew Marantz lays out these episodes in roughly 18 000 words, drilling into every twist: secret handshakes, war-room strategy, and board complaints about his pattern of deception.
Elon Musk’s split from OpenAI came down to who controls the C-suite. Musk claimed he left over safety and governance concerns, but all three founders—Altman, Brockman and Sutskever—say Musk’s version doesn’t match their secret pact. They’d agreed that Altman would step down if both Brockman and Sutskever asked. When the board learned he’d essentially created his own shadow leadership team, they felt blindsided. Meanwhile, Dario Amodei’s departure hinges on a last-minute charter change. He’d demanded a “merge-and-assist” clause be locked in for safety reasons. But as Microsoft’s investment closed, a merger-blocking provision slipped into the charter without his knowledge. Altman denied it was there until Amodei forced him to read it aloud. That breach of trust sealed the break.
Behind the scenes, Altman pushed Summers and Taylor to kill off their promised investigation into his actions. They narrowed their scope to potential criminality, issued no public report and gave Altman a clean bill. Current and former employees say many board members never saw any briefing. Meanwhile, the article paints Altman as candid on some points—like decision fatigue—but evasive on others. It points out how every retelling shifts according to who’s talking: Altman, Musk, Amodei or board members. You come away with a clear sense that OpenAI’s biggest risks haven’t been purely technical—they’ve been about trust, control and conflicting incentives.