OpenAI, the company leading the artificial intelligence boom and the maker of ChatGPT, had a rather tumultuous weekend.
The company has now had three CEOs in as many days, and its high-profile leader, cofounder and former CEO Sam Altman, has decamped to start a formidable rival embedded at Microsoft. Now even that seems in doubt, with some indications that Altman is still holding out hope for a return to OpenAI.
Altman’s departure puts OpenAI’s $86 billion valuation at risk, which would hurt investors, executives, and employees with a stake in the company, right as they were gearing up to enter a financial agreement that would have allowed them to sell portions of their highly valued shares.
And the events of this weekend underscore a larger debate that’s been going on since AI burst into the zeitgeist at the beginning of the year over how to develop a potentially transformational technology safely and ethically. Critics are wary that Silicon Valley’s “move fast and break things” ethos could have drastic consequences when it comes to a technology that some see as an existential threat to humans.
Give me the TL;DR: What happened at OpenAI this weekend?
On Friday afternoon, OpenAI’s board removed Altman from his position, setting off feverish speculation in the tech world with a blog post saying that he hadn’t been “consistently candid in his communications.” The board didn’t elaborate on what he had not been candid about. The company’s CTO, Mira Murati was appointed interim CEO.
There had been rumors of increasing tensions between Altman and his cofounder Ilya Sutskever, who had a seat on OpenAI’s board. Altman’s fellow cofounder, OpenAI president Greg Brockman, was also removed from his role as board chair. Several hours later, Brockman would resign from his role in solidarity with Altman.
By Saturday it was reported that Murati was in talks to hire Altman and Brockman back at the company in different capacities. Altman seemed open to that possibility, even posting a picture of himself in the OpenAI office’s wearing a guest badge. “First and last time I wear one of these,” Altman wrote on X. The attempt to bring Altman back came after mass employee protests and Altman’s allies, including Microsoft executives, pushing for his return.
At this point, it became clear OpenAI wouldn’t allow Altman to return to the company. By early Monday morning, Microsoft CEO Satya Nadella had swooped in to hire Altman and Brockman to lead a newly founded research arm within the tech giant.
Who is Sam Altman, anyhow?
Altman has become a major figure in Silicon Valley and by far the most recognizable face in AI. Some have compared his ouster to Apple’s famously misguided firing of a young Steve Jobs.
Prior to working full time at OpenAI, Altman was the president of Y Combinator, the most renowned startup incubator in Silicon Valley. Altman’s own startup, Loopt, which lets users share their locations with each other, was a member of the first class of startups incubated at Y Combinator.
What did the board and Altman disagree over?
That’s still unclear, even to tech insiders. While there hasn’t been any definitive answer, several possibilities have been raised. Given that Altman is a serial entrepreneur, it is possible the board chafed at some of his current or planned business ventures, which might give the impression of a possible conflict of interest.
Altman had been in the Middle East looking for backers to start a new chip company to rival Nvidia, the dominant chip-maker in AI, according to Bloomberg. The rise of AI has led to a boom in demand for computing power.
Another possibility is that philosophical disagreements between Altman and Brockman and the board came to a head. Some board members with ties to the effective altruism movement (which disgraced crypto executive Sam Bankman-Fried was also a proponent of) believe AI poses a grave threat to humanity. Subscribers to that school of thought are wary of advancing AI just for the sake of technological development, while those in Altman’s camp wanted to push OpenAI’s research as far along as it could go. The board may have considered Altman and his allies too gung-ho about pushing forward the technology without properly considering the potential dangers.
Why did this happen so suddenly?
Well, in truth it didn’t. This weekend’s drama followed a year’s worth of tensions and concerns from Sutskever and his allies over Altman’s increasingly rapid push to commercialize OpenAI’s artificial intelligence products.
At OpenAI’s first developer day earlier this month, Altman announced a tool that would let other developers use its models to make their own AI tools. Some executives at the company, including Sutskever, were uncomfortable with this idea, according to CNN.
What’s Microsoft’s role in all this?
Microsoft is OpenAI’s biggest financial backer and provider of the computing power needed to run its powerful models, having invested a total of $13 billion in the company, so it has an unusually large influence in the proceedings. Executives from Microsoft, including Nadella, learned of OpenAI’s decision to fire Altman just one minute before it was announced publicly, according to Axios.
At first, Nadella said Microsoft remained committed to its partnership with OpenAI even though Altman would no longer be helming the company. On Saturday, however, Microsoft joined other investors in pushing for Altman to return to OpenAI—and eventually hired him and Brockman.
How do OpenAI’s employees feel about this?
In one word: mutinous. On Monday morning, over 500 of OpenAI’s 700 employees signed a letter saying they would quit and join Altman at Microsoft if the board didn’t resign.
Altman was a popular CEO during his time at OpenAI, and employees were shocked by the news of his departure, which they only found out about when the announcement was made public outside the company.
OpenAI was about to sign a “tender offer” that would have let employees sell some of their future profit participation rights to other investors. Altman’s departure puts that agreement in jeopardy, and could cost those employees the chance of a spectacular payout.
The letter criticized the board’s handling of Altman and Brockman’s removals did not mince words: “Your conduct has made it clear you did not have the competence to oversee OpenAI,” it reads. Murati was the first signatory, and Sutskever, the board member who disagreed with Altman in the lead up to his firing, also signed.
What could happen next?
For Microsoft the plan is relatively straightforward: It will funnel resources—money, personnel, and cloud computing power—into the new research arm Altman will lead. Altman will continue the same work he was doing at OpenAI, building increasingly sophisticated large language models and ideating commercial uses for them—perhaps with a staff including defectors from OpenAI.
Things at OpenAI seem much more tumultuous, as the threat of either a mass employee resignation or a total overhaul of the board—or both—looms. The company also faces the possibility of mass defections to Microsoft. In the meantime, Shear announced plans to investigate the circumstances of Altman’s firing.
And it will have to contend with disgruntled investors upset that OpenAI’s key man has departed for a rival that also happens to be one of tech’s biggest firms. If OpenAI ends up losing its lead in the AI race to Microsoft, then its investors may watch the value of their investments be wiped out. (And of course, Microsoft is one of these investors.)
As for Altman, free from the shackles of OpenAI’s unusual corporate structure as a for-profit subsidiary of a non-profit, he can pursue any and all paths to advance AI. Some have wondered, however, whether the longtime startup exec will find the rigidity of a corporate behemoth like Microsoft constricting. Meanwhile, sources told The Verge that Altman may still be angling for a return to OpenAI.
What does this mean for the future of AI?
Regardless of the various corporate machinations that occur, the weekend’s events illustrate the high-stakes consequences of artificial intelligence research.
This is not your standard tech company personnel drama: Unlike dating, ride sharing, or food delivery apps, AI’s future is tied to the future of our species, as the author Yuval Noah Harari has written. “Potentially we are talking about the end of human history,” Harari has warned, “the end of the period dominated by human beings.”