The New Yorker has uncovered the truth behind OpenAI's palace coup, with 100 interviews and 70 pages of secret memos. The truth behind Sam Altman's reinstatement has finally been revealed!!! The board of directors may not have been crazy back then. This report is a re compilation of the controversy using over 100 interviews, approximately 70 pages of secret memos from Ilya, and over 200 pages of internal documents related to Dario Amodei. The most eye-catching sentence is the first behavior pattern listed by Ilya for Sam in the memo, which is directly: Lying. What's even more outrageous is that OpenAI, while boasting about security, has been exposed for promising 20% computing power to the Super Alignment team, but in reality only giving 1% -2%, and many of them are even the oldest clusters. The board later discovered that some of the most sensitive features of GPT-4 had not been approved for security as originally agreed. Microsoft launched early ChatGPT in India without completing the necessary review. So looking back now, the recall in 2023 was not just a 'board fight'. It was more like a governance mechanism trying to hit the brakes, only to be pushed back forcefully by capital, employee turnover risks, and Sam's control over the power structure. Because what truly determines the outcome is not whether there are problems, but who can better mobilize resources. Thrive has put the $86 billion valuation financing on hold, and employee cash out is tied to Sam's return. Sam even sent a text message directly to Nadella, even thinking about the framework of 'who will investigate me' first. The most alarming thing is not actually the palace intrigue itself. But rather OpenAI's earliest story of "having to use a special governance structure to constrain itself in order to prevent AGI from losing control" has been largely washed away by reality. An organization that originally claimed to be "for the safety of all mankind" is now moving towards a potential trillion yuan IPO while increasingly embedding its business into government and defense scenarios. Ultimately, OpenAI's biggest shift in recent years may not be from non-profit to for profit. It has gradually shifted from "managing risks first" to "stabilizing power, computing power, and orders first".