The truth about the creator of ChatGPT

The New Yorker's investigation, based on approximately 200 pages of internal documents and over 100 interviews, concluded that Sam Altman consistently lied and prioritized power over AI safety.

In November 2023, the OpenAI board of directors unexpectedly fired Sam Altman. Just five days later, he was reinstated and immediately replaced the old board with a group of close associates. OpenAI employees dubbed this event "The Blip," inspired by an event in a Marvel film where a character disappears and then returns while the world around them has undergone profound changes.

In a lengthy investigation by The New Yorker, based on 200 pages of previously unpublished internal documents and more than 100 interviews, what happened in those five days was not simply a power struggle. It was the culmination of years of accumulated concerns about the man who controlled the world's most powerful AI company.

The record is a lie.

In the fall of 2023, Ilya Sutskever, OpenAI's chief scientist , sent self-deleting memos to three board members. He compiled approximately 70 pages of Slack messages and personnel documents, photographing them with his phone to avoid detection on company devices. One of the memos began with a list alleging Sam Altman was lying.

Dario Amodei, who led OpenAI's security division before leaving to found Anthropic, also meticulously documented his activities for many years. His more than 200-page document describes a specific series of deceptive behaviors by the company's CEO. In 2019, during negotiations for an investment from Microsoft, Amodei provided Altman with a list of security requirements.

Sam Altman failed to fulfill OpenAI's core commitment. Photo: Bloomberg .

As the contract was about to be signed, a clause excluding the most important requirement emerged. When Amodei confronted him, Altman denied the clause existed, even when Amodei read it aloud to him.

"OpenAI's problem is Sam," Amodei wrote in the note.

"Sam possesses two traits rarely seen in one person. First, an intense desire to please others in every interaction. Second, an almost pathological indifference to the potential consequences of deceiving others," a former OpenAI board member told the New Yorker .

Regarding AI security, a core commitment when OpenAI was founded, it was never given enough attention. In 2023, the artificial intelligence giant pledged to dedicate 20% of its computing power to its security research team. In reality, the team only received 1-2% and was soon disbanded due to incompatibility with the company's strategy.

5 days of chaos

Altman was in Las Vegas watching Formula 1 racing when he received a video call from the board of directors. At that moment, Sutskever called Altman and said that he was no longer the CEO of OpenAI, citing a lack of communication.

Satya Nadella, CEO of Microsoft, was not informed of this beforehand. "I was stunned. I couldn't get the information from anyone," Nadella said.

Within hours, Altman had set up what he called "an exile team" at his $27 million San Francisco mansion. Investor Josh Kushner suspended the $86 billion deal and stated it would only be completed if Altman returned. Microsoft announced its readiness to create a competing initiative and recruit employees who left OpenAI following the event.

Sam Altman returned to OpenAI as a very different person after being fired. Photo: Bloomberg .

A letter requesting Altman's return circulated within the company. Eventually, a majority of employees threatened to quit, leaving the board with no choice. The upheaval was swift, resulting in Sutskever, Toner, and McCauley losing their board seats. Two new members, former Harvard president Larry Summers and former Facebook CTO Bret Taylor, were chosen by Altman after private conversations with him.

The subsequent investigation by WilmerHale did not produce a written report. Insiders told the New Yorker that the investigation focused solely on finding clear evidence of wrongdoing and concluded that Sam Altman could continue as CEO. "The review did not conclude that Sam was a person of perfect integrity," said a person close to the investigation.

Unlimited power

Since The Blip event, OpenAI has become one of the world's most valuable companies, preparing for an IPO with a potential valuation of up to $1 trillion . Altman is one of the entrepreneurs most favored by the Trump administration, accompanying the president on his visit to the British Royal Family and sitting across from Trump at dinner at the White House.

Sam Altman was favored under President Trump. Photo: Bloomberg .

In January 2025, Altman stood in the Roosevelt Room announcing Stargate, a $500 billion venture to build AI infrastructure across the United States, with plans to expand to the UAE. Just before that, Trump repealed the entire Biden administration's AI safety executive order, which Altman had helped draft and once praised as "a good start."

Altman's journey reflects a perplexing paradox. OpenAI was founded on a non-profit commitment, prioritizing human safety over revenue. But according to former leaders, that commitment ultimately succumbed to commercial pressures.

"He established structures that, in theory, would restrict him in the future," said Carroll Wainwright, a former OpenAI researcher.

Former board member Sue Yoon offered the most balanced assessment, stating that Altman wasn't a scheming villain. Instead, the OpenAI CEO was someone too consumed by his own beliefs. "He does things that wouldn't make sense in the real world," Sue Yoon said.

Source: https://znews.vn/su-that-ve-con-nguoi-sam-altman-post1641356.html