Behind the Scenes of Sam Altman’s Wild Ouster From OpenAI
A fired CEO, middle-finger emojis and the battle royale over the future of artificial intelligence
Sam Altman was in Las Vegas for the biggest party of the weekend, the Formula One race, when he opened the Google Meet link that would set off a whiplash-inducing corporate power struggle about the future of artificial intelligence.
In the moments before he clicked on the link in his hotel room, the 38-year-old CEO was living a charmed life. Ever since his company, OpenAI, ushered in the AI age a year ago with the release of ChatGPT, Altman had been jet setting around the world meeting with presidents and prime ministers, shaping the global agenda on AI policy. And he was nearing a fundraising that would value OpenAI at nearly $90 billion, roughly triple what it had been earlier in the year.
At noon, he logged on to find the company’s board—except, ominously, his closest ally and co-founder, Greg Brockman—peering back at him. Ilya Sutskever, the brilliant, eccentric AI researcher whose hiring eight years ago had put OpenAI on the map, informed Altman that he was being fired and that the news would be announced soon. No explanation was given.
And then, shortly after the little boxes blinked closed, Altman was locked out of his computer.
The crisis at OpenAI is personifying a question that has been boiling inside the AI industry and creating angst among technology giants and world leaders: Who can be trusted to open the Pandora’s box that artificial intelligence might represent?
One solution that Altman devised was a curious corporate structure that led to his ouster. A nonprofit board governs OpenAI’s for-profit business arm with the sole purpose of ensuring the company develops AI for humanity’s benefit—even if that means wiping out its investors.
It was this board that abruptly fired Altman, even as OpenAI’s business hit its apex. The board has been vague in public about its reasons, saying in a blog post that Altman was out because he “was not consistently candid in his communications with the board.” According to people familiar with the board’s thinking, members had grown so untrusting of Altman that they felt it necessary to double check nearly everything he told them.
The explanation baffles Altman’s defenders, who say they aren’t aware of specific episodes that might warrant such an outcome.
Over the weekend, Altman’s old executive team pushed the board to reinstate him—telling directors that their actions could trigger the company’s collapse.
“That would actually be consistent with the mission,” replied board member Helen Toner, a director at a Washington policy research organization who joined the board two years ago.
This article is based on interviews with more than a dozen insiders at OpenAI and people around the company’s hectic weekend.
The situation remains fluid, including where Altman will work or how many followers he would take if he left, and discussions are continuing between the camps. The near-$90 billion-valuation investment plan is on hold.
Altman and Elon Musk had set out to create OpenAI in 2015 as a nonprofit aimed at achieving artificial general intelligence, or AGI, a system with reasoning capabilities that match or exceed a human’s. They declared they wanted to do this in a way that would benefit humanity, not just to make profit for corporations. But it turned out that the most promising technological path they found required vast stores of computational power, and thus great piles of money, to work.
Despite the “weird,” in his own parlance, nonprofit structure, Altman managed to convince much of the biggest money in the Valley to sign on, including Khosla Ventures and Founders Fund. Eventually, he reeled in
, which invested $13 billion for 49% of the for-profit.OpenAI’s leaders had a long history of bitter feuds. There was the rupture between Altman and Musk, who left in 2018. Then there was another large rift a couple of years later, when a group of key researchers left to found the rival company Anthropic after clashes over safety.
This time around, Altman’s grip on the board slipped after some of the more business-minded board members left earlier this year. The maker of the most advanced AI technology that was rapidly weaving itself into every nook and cranny of the American economy came to be controlled by four people who weren’t focused on whether the business was economically successful.
In addition to Sutskever, OpenAI’s board consists of Adam D’Angelo, a former Facebook executive and a co-founder of the question-and-answer website Quora; Tasha McCauley, an adjunct senior management scientist at the Rand Corp., a policy nonprofit, and Toner, a director at the Center for Security and Emerging Technology, a research organization tied to Georgetown University. The center receives funds from Open Philanthropy, a group identified with effective altruism, a philanthropic movement whose adherents spend money trying to solve potentially catastrophic problems affecting the world.
Friday, at noon, those four people fired Altman. At 12:19 p.m., with Sutskever as their spokesman, they removed Brockman from the board.
Brockman resigned from his employee role in protest.
Inside OpenAI’s plant- and fountain-filled offices in San Francisco, the executive leadership team was in shock. The board had named Chief Technology Officer Mira Murati interim CEO and handed the executive team a packet of crisis-communications talking points that offered no more insight into what was going on than the board’s vague blog post, according to people familiar with the matter.
Murati and Sutskever led an all-hands meeting at 2 p.m. Employees peppered them with dozens of questions, many of which were some version of: what did Sam do? One employee asked if they would ever find out, to which Sutskever replied, “No.”
After that meeting, the executive team regrouped in a conference room. A member of the executive team told Sutskever that the lack of detail was unacceptable and demanded the rest of the board join a video call to explain, according to people familiar with the matter.
On the call, the leadership team pressed the board over the course of about 40 minutes for specific examples of Altman’s lack of candor, the people said. The board refused, citing legal reasons, the people said.
Some executives said they were getting questions from regulators and law-enforcement entities such as the U.S. attorney’s office in Manhattan over the charge of Altman’s alleged lack of candor, the people said. The truth was going to come out one way or another, they told the board.
People familiar with the board’s thinking said there wasn’t one incident that led to their decision to eject Altman, but a consistent, slow erosion of trust over time that made them increasingly uneasy. Also complicating matters were Altman’s mounting list of outside AI-related ventures, which raised questions for the board about how OpenAI’s technology or intellectual property could be used.
The board agreed to discuss the matter with their counsel. After a few hours, they returned, still unwilling to provide specifics. They said that Altman wasn’t candid, and often got his way. The board said that Altman had been so deft they couldn’t even give a specific example, according to the people familiar with the executives.
The executives requested written examples of the board’s allegations.
Meanwhile, Altman was on the phone with Satya Nadella, the CEO of Microsoft, saying he wanted to keep working on the technology. They gamed out ways to undo the day’s events, but also began to hash out a backup plan for Altman to bring a bunch of his top researchers and start a new division at the tech giant, according to people familiar with their conversation.
Altman also told friends that he was thinking of starting a new company with Brockman and intended to hire away dozens of OpenAI employees.
Altman blamed himself for not better managing the board, which he felt was taken over by people overly concerned with safety and influenced by effective altruism.
The specter of effective altruism had loomed over the politics of the board and company in recent months, particularly after the movement’s most famous adherent, Sam Bankman-Fried, the founder of FTX, was found guilty of fraud in a highly public trial.
Some of those fears centered on Toner, who previously worked at Open Philanthropy. In October, she published an academic paper touting the safety practices of OpenAI’s competitor, Anthropic, which didn’t release its own AI tool until ChatGPT’s emergence.
“By delaying the release of Claude until another company put out a similarly capable product, Anthropic was showing its willingness to avoid exactly the kind of frantic corner-cutting that the release of ChatGPT appeared to spur,” she and her co-authors wrote in the paper.
Altman confronted her, saying she had harmed the company, according to people familiar with the matter. Toner told the board that she wished she had phrased things better in her writing, explaining that she was writing for an academic audience and didn’t expect a wider public one. Some OpenAI executives told her that everything relating to their company makes its way into the press.
OpenAI leadership and employees were growing increasingly concerned about being painted in the press as “a bunch of effective altruists,” as one of them put it. Two days before Altman’s ouster, they were discussing these concerns on a Slack channel, which included Sutskever. One senior executive wrote that the company needed to “uplevel” its “independence”—meaning create more distance between itself and the EA movement.
OpenAI had lost three board members over the past year, most notably Reid Hoffman, the LinkedIn co-founder and OpenAI investor who had sold his company to Microsoft and been a key backer of the plan to create a for-profit subsidiary. Other departures were Shivon Zilis, an executive at Neuralink, and Will Hurd, a former Texas congressman.
The departures left the board tipped toward academics and outsiders less loyal to Altman and his vision.
Friday, after being fired, Altman immediately flew back to San Francisco. On Saturday, his Russian Hill home became a war room filled with OpenAI employees, including Murati, then the interim CEO, and other members of the executive team, plotting his return to the company.
They began to use X in a coordinated fashion for their campaign. That evening, Altman tweeted “i love openai employees so much,” and dozens of OpenAI employees quote-tweeted it with heart emojis.
On Sunday morning, Murati sent a note to staff saying that Altman would be returning to the office. Altman arrived, mugging on X with his guest pass, writing “first and the last time I ever wear one of these.”
He, Brockman, Murati, strategy chief Jason Kwon, Chief Operating Officer Brad Lightcap and the rest of the leadership team spent the day negotiating with the board, arguing for Altman’s reinstatement and tossing out suggestions for new board members. They believed they were getting somewhere. One suggestion, Bret Taylor, the former co-CEO of
, was approved by both sides.But the sticking point remained Altman’s desire to reclaim his CEO role, people familiar with the matter said. Altman also pushed to fire the current board.
The negotiations stretched late into the night, as multiple rounds of takeout were delivered to the Mission District offices and reporters from multiple news outlets staked out the scene like a papal conclave. Inside, employees gathered, some openly sobbing.
When the white smoke came, it wasn’t Altman. Emmett Shear, the co-founder of video-streaming service Twitch and a vocal proponent of a slower approach to developing AI, had been named CEO by the board.
Microsoft was given no heads-up about the decision, but when it dropped Nadella had a plan at the ready: Altman and Brockman would go to Microsoft to start a new AI unit, one person said.
The board then put out a memo to the OpenAI team saying it stood by its decision and thought Altman’s “behavior and lack of transparency” had “undermined the board’s ability to effectively supervise the company in the manner it was mandated to do.”
In an attempt to bat down the theories coursing through social media, it added that “This decision is not about product safety or security, the pace of development or OpenAI’s finances.”
“Bottom line,” the board said, “This was a governance issue that lies at the heart of how the board of this uniquely structured organization executes its responsibilities and advances its mission.”
A Slack message went out to employees announcing an all-hands meeting with Shear, but several employees responded with middle finger emojis, according to a person familiar with the matter. Less than a dozen people showed up. The roughly 200 people who had spent their Sunday waiting for a resolution flooded out of the building.
In the lobby, Brockman’s wife, Anna Brockman, who had been married in 2019 at OpenAI’s offices in a civil ceremony officiated by Sutskever, cried and pleaded with Sutskever to reconsider.
Shortly afterward, he did.
Overnight Sunday night, OpenAI employees penned a blistering open letter threatening to quit and follow Altman and Brockman to Microsoft unless the board resigned and reinstated them, and installed new independent board members like Taylor and Hurd. By the end of the day, more than 700 of the 770 employees had signed it—including Sutskever.
“I deeply regret my participation in the board’s actions,” he wrote on 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.” Altman retweeted it with three hearts.
On Monday evening, Nadella went on CNBC and said OpenAI needed a new governance structure, but he left the door open to the possibility that Altman and his team would remain at OpenAI.
Microsoft has set up a floor of the LinkedIn office in San Francisco with laptops and clusters of GPU chips, which provide the computational power behind AI, to house the OpenAI team if the bid to restore Altman to the company he co-founded fails.
But as negotiations drag on between the board and OpenAI leadership, there are no OpenAI employees there yet. A person familiar with Nadella’s thinking said Microsoft’s first preference is for Altman to return as OpenAI CEO.
Corinne Ramey and Tom Dotan contributed to this article.
Write to Keach Hagey at Keach.Hagey@wsj.com, Deepa Seetharaman at deepa.seetharaman@wsj.com and Berber Jin at berber.jin@wsj.com
'New Vocabularies' 카테고리의 다른 글
| Mystery of the Polar Bear Intrusion in Langjökull Solved (0) | 2023.12.14 |
|---|---|
| Australia's Move to Curb Youth Vaping (0) | 2023.12.14 |
| Israel Says 80,000 People Left Northern Gaza Thursday - 11/10/2023 (0) | 2023.11.10 |
| Mexican Sinaloa Cartel’s Message to Members: Stop Making Fentanyl or Die - 10/19/2023 (0) | 2023.10.20 |
| Don’t Hire My Anti-Semitic Law Students - 10/18/2023 (0) | 2023.10.19 |