Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
OpenAI’s directors are in talks with Sam Altman to allow him to rejoin the board, four days after their decision to sack him plunged the generative artificial intelligence start-up into turmoil.
A deal to unify the company by bringing its former chief executive back alongside the remaining directors would be a compromise for both sides. More than 95 per cent of OpenAI employees signed a letter this week calling for the board to resign and reinstate Altman, while a trio of holdout directors have remained resolute in their view that his firing was justified.
The option, first reported by Bloomberg, is one of a number being discussed by the non-profit board that ultimately controls OpenAI, which spectacularly removed Altman and his co-founder Greg Brockman as directors last week, according to people with direct knowledge of the negotiations. After being stripped of his role as chair of the board, Brockman quit the company on Friday.
Staff at the company — led by executives Mira Murati, Brad Lightcap and Jason Wong — have thrown their weight behind the co-founders and pressed the board for more detailed answers as to why Altman was fired. But as of Tuesday afternoon, the sides had not reached an agreement on the company’s future.
The departures of Altman and Brockman triggered a chaotic few days at OpenAI, which has become the most celebrated start-up in Silicon Valley since it launched its ChatGPT chatbot a year ago, kicking off a boom in generative AI.
Ilya Sutskever, a third co-founder, was one of the four directors who voted to oust Altman. Under growing pressure from his colleagues, Sutskever signed the letter calling on the board to reverse course and apologised on social media on Monday.
“I deeply regret my participation in the board’s actions,” he wrote on social media platform X. “I never intended to harm OpenAI. I love everything we’ve built together and I will do everything I can to reunite the company.” Sutskever did not say whether he would step down from the board.
That left three directors opposed to Altman’s return: Adam D’Angelo, chief executive of question-and-answer service Quora; technology entrepreneur Tasha McCauley; and Helen Toner from the Center for Security and Emerging Technology at Georgetown University.
The trio have come under growing pressure from employees and investors in OpenAI’s for-profit entity to explain their decision and reverse course. Toner was described by two investors in the company as the hardest of the board members to bring onside.
Toner co-authored an academic paper that compared the approaches to safety taken by OpenAI and rival company Anthropic. Its publication in October triggered a clash between her and Altman, according to the New York Times.
OpenAI’s launch of ChatGPT in November 2022 created “race to-the-bottom dynamics” as rivals scrambled to compete, argued the paper, which also noted “safety and ethics issues related to the launches of ChatGPT and GPT-4”.
In contrast, Anthropic’s decision to delay the release of its own chatbot avoided “stoking the flames of AI hype . . . Anthropic enhanced the credibility of its commitments to AI safety by holding its model back from early release”, said the paper.
Before Altman was fired, questions had been raised internally around whether the pace of AI development at the company was safe and about potential conflicts with the 38-year-old entrepreneur’s side projects, which range from cryptocurrency to nuclear fission. The board had also lost trust in Altman, according to a person with knowledge of their thinking.
But investors in the company said they had been left in the dark as to the specific reason for his firing.
Emmett Shear, the co-founder of video-streaming service Twitch whom the board appointed as interim chief executive on Sunday, has also called for an independent investigation into how the decision was made to oust Altman and pledged to reform the management of the company.
Investors, including Microsoft, would like to see governance changes to protect OpenAI from a similar crisis in future, according to multiple people with direct knowledge of their thinking. Those changes will not be discussed until there is clarity on the narrow question of Altman’s future at the company, said one of the people, adding that until then, “there’s no one to negotiate with”.
Read the full article here