In an interview with the TED AI Show published Tuesday afternoon, former OpenAI board member Helen Toner revealed new details about the company's efforts to fire billionaire CEO and co-founder Sam Altman, and why Toner “never believed” anything he said.
November 2023, The board of directors of OpenAI, the developer of ChatGPT, decided that Altman had “not been consistently forthright in his communications” and fired him. A few days after the firing, 95% of the company's employees signed a letter of warning threatening to resign if Altman was not reinstated.
Altman returned to his job less than a week after being fired and remains CEO of OpenAI, but one question remains after the attempted firing: Why did the board try to fire him in the first place?
Related: Scarlett Johansson is 'shocked' that OpenAI used a voice that was 'eerie' like hers after telling her 'no'
On the podcast, Toner, the director of research and strategy at Georgetown University, said board members have reached a point where they can't trust Altman.
Toner alleged for years that Altman “concealed information and misrepresented what was going on at the company” and “outright lied to the board.”
Helen Toner. Jerrod Harris/Getty Images, courtesy of Vox Media
Toner gave specific examples, first stating that when ChatGPT was released in November 2022, the board of directors learned of the release through Twitter, but did not know it would be released in advance.
What's Really Going on at OpenAI? Former Director Follow host translation: For a behind-the-scenes look at the future of government regulation of AI, watch the TED AI show. Amazon Music Or listen to the podcast: https://t.co/wZh1JsVy7m pic.twitter.com/kc31ElaZiC
— TED Talks (@TEDTalks) May 28, 2024
Toner said Altman also failed to tell the board that he “owned” the OpenAI Startup Fund, a $175 million fund for early-stage AI companies. Toner said Altman's ownership was inconsistent with his claim that he was an “independent director with no financial interest in the company.”
The fund's website states that at the time of writing, “OpenAI itself is not an investor.”
Toner also accused Altman of providing the board with “inaccurate information about the few formal safety measures the company actually had in place” and that the board did not know how well the company's AI safety features were performing or what changes were needed.
Two of OpenAI's top safety researchers recently left the company, and OpenAI disbanded the groups they led and formed a new safety team this week, led by Altman.
Related: Now that OpenAI's superalignment team has been disbanded, who will stop AI from running wild?
Two OpenAI executives also told the board they did not believe Altman was the right person to lead the company, and shared screenshots and documents that exposed a “toxic” atmosphere.
“All four of us who fired him came to the conclusion that we couldn't believe what Sam was saying,” Toner said. “As a board, this is a totally unworkable situation, especially as a board that is supposed to provide independent oversight to the company and not just help the CEO raise more money.”
When asked why 95% of OpenAI's staff wanted Altman back, Toner said employees may have been told either he would return or the company would be destroyed.
OpenAI responded to Toner's comments by saying that it conducted a “thorough investigation” into the board's attempt to fire Altman and found that the decision was “not based on concerns about the safety or security of our products, the pace of our development, OpenAI's finances, or statements to our investors, customers or business partners.”
The Giving Pledge announced Tuesday that Altman has pledged to donate the majority of his fortune.
