After more than six months Sam Altman was fired, then rehired Finally, one of OpenAI’s former board members is spilling the tea on what happened behind closed doors. Helen Toner, one of the four people responsible for firing the OpenAI CEO, says Altman’s constant lying created a toxic culture that executives described as “psychological abuse.”
Toner says in her first lengthy interview since Sam Altman’s firing ted presentation artificial intelligence That executives came to OpenAI’s board in October 2023 with serious allegations against the company’s CEO. According to Toner, two executives said they could not trust Altman and showed the board screenshots of Altman’s manipulation and lying. These executives reportedly said they did not believe Altman could or would change, and their testimony prompted the board to fire the CEO weeks later. This interview, released on Tuesday, comes after weeks of public backlash against OpenAI where the company’s credibility has been compromised It has been questioned by Scarlett Johansson and former employees.
“For any individual case, Sam could always come up with some kind of harmless explanation for why it didn’t matter or why it was misinterpreted or whatever,” Toner said in the interview. “After years of this kind of thing, all four of us who kicked him out came to the conclusion that we couldn’t believe the things Sam was telling us.”
But the writing has been on the wall about Altman’s alleged lying for years, according to Toner. It says the board was not informed in advance when ChatGPT was released in November 2022, and “learned about ChatGPT on Twitter.” Toner also noted that Altman provided inaccurate information about OpenAI’s safety processes. In the weeks before Altman’s firing, Toner claimed he lied to other board members to try to fire her after she wrote a research paper that spoke negatively about OpenAI’s safety practices.
In the end, Toner says, OpenAI board members didn’t tell anyone except its legal team that they would try to fire Altman because they knew the CEO would try to undermine them if he found out. But even after all this, Altman He returned as CEO just a few days laterwith 95% of the company signing an open letter to reinstate him.
Toner says this was presented as a black-and-white decision to employees within the company: either bring back Altman or destroy OpenAI. The company’s security and valuation were especially important, according to Toner, because OpenAI employees would make a lot of money from their shares in the $86 billion company via a tender offer a few months later.
“The second thing that’s really important to know, and that’s not been reported enough, is how afraid people are of confronting Sam,” Toner said. “They experienced him taking revenge on people, taking revenge on them, for his criticisms in the past. “They were really afraid of what might happen to them.”
Finally, Toner noted that this is not the first company where Altman has encountered this problem. The former OpenAI board member noted that Altman was fired from Y Combinator in 2019, which… Washington Post It was reported in the wake of his firing from OpenAI. Toner also said that the management team at Loopt, Altman’s first startup, went to the company’s board twice and asked them to fire Altman for “deceptive and disruptive behavior.”
Toner, Tasha McCauley, Ilya Sutskever and Adam D’Angelo were the board members responsible for firing Sam Altman last November. Toner and McCauley immediately left the OpenAI board when Altman returned to power later that month. Sutskever just announced He passed away this monthafter he was reportedly absent from the OpenAI office for approximately six months.
In response to this wave of allegations, the podcast included a response from OpenAI CEO Brett Taylor. “We are disappointed that Ms. Toner continues to revisit these cases,” Taylor said, then cited the independent investigation conducted by the law firm Wilmer Hill into the cases. “The review concluded that the Board’s previous decision was not based on concerns regarding the safety or security of the product, the pace of development, OpenAI’s finances, or its data to investors, customers, or business partners.”
This interview comes after weeks of turmoil for OpenAI, as the company’s credibility began to become increasingly apparent to the public. OpenAI has also been criticized for… Strict exit contracts that muzzle former employees They threaten to take back their shares (the company withdrew these contracts in light of the public backlash). Finally, it saw OpenAI The departure of several high-level AI safety researchersMany issued a warning call about the company when they left. Six months after the Altman launch debacle, OpenAI’s trust issues don’t appear to be going away anytime soon.
“Infuriatingly humble alcohol fanatic. Unapologetic beer practitioner. Analyst.”