After an enormous waste of capital, time and trust, Sam Altman and Greg Brockman return to OpenAI. The board is renewed.
After being fired last Friday and presumably leaving for Microsoft, Sam Altman has returned to OpenAI as CEO. He is followed by Greg Brockman, whose new role is still unclear.
The incident will at least have an impact on the structure of OpenAI: The new board consists of Bret Taylor, former co-CEO of Salesforce, former Treasury Secretary Larry Summers and Adam D’Angelo, CEO of Qora, who was the only one of those involved to survive the incident in his position. Altman and Brockman will not return to the board.
According to The Verge, this initial board will form an expanded board of up to nine people, including Microsoft. For now, Altman, Brockman, Sutskever, Tasha McCauley and Helen Toner are no longer on the board. The latter is said to have been heavily involved in the disputes of the past few days (see below).
Interim CEO Emett Shear, who helped with the discussions, is likely to leave the company now that Altman is back. On Twitter, he says the solution now is “the path that maximizes safety while doing right by all stakeholders.”
OpenAI embarrasses its partner Microsoft
Microsoft CEO Satya Nadella is likely to be quite upset by what has happened, as he ended up embarrassing himself by prematurely announcing that he had brought Altman and Brockman to Microsoft.
Nadella had already been hailed by many as the winner of the OpenAI conspiracy. On Monday, many Microsoft employees took to LinkedIn to welcome the new AI research team with Altman and Brockman at the helm.
“When I decided to join Microsoft on sun evening, it was clear that was the best path for me and the team. With the new board and with Satya’s support, I’m looking forward to returning to OpenAI, and building on our strong partnership with Microsoft,” said Altman, explaining his change of heart.
More likely, Altman wanted to use his move to Microsoft and the potential hiring of hundreds of OpenAI employees as leverage in negotiations with OpenAI’s board, as well as his threat to start a new company with Brockman.
Nadella credits Altman and Brockman with playing key roles in OpenAI and says he welcomes the change in the board. “We look forward to building on our strong partnership and delivering the value of this next generation of AI to our customers and partners.”
Dispute over AI safety and a research paper
It is still unclear exactly what triggered the dispute, but it seems to be a mixture of reasons. According to the New York Times, Ilya Sutskever is said to have expressed concerns about the rapid commercialization being pushed by Altman and the safety risks involved.
Sutskever had concerns about AI safety and Altman’s role since the Anthropic spin-off. Tasha McCauley also had these concerns, even more so than Sutskever, and had previously clashed with Altman, Kara Swisher reported on Twitter.
Altman also reportedly wanted to remove Helen Toner from the board for co-authoring a research paper that criticized OpenAI’s safety measures while praising OpenAI spin-off and competitor Anthropic. The New York Times saw an email in which Altman expressed these concerns.
Altman reportedly confronted Toner about the paper. He saw it as a risk to OpenAI, given that in July the FTC launched an investigation into ChatGPT for spreading misinformation about individuals and OpenAI’s handling of user data. When Sutskever sided with Toner, events took their course.