It’s official: OpenAI has named a new board of directors — three men: Bret Taylor, the chair of the board and the president and chief operating officer of Salesforce; Larry Summers, a former U.S. Treasury Secretary and a professor at Harvard University; and Adam D’Angelo, co-founder and CEO of Quora/Poe and the only carryover from the prior board, oh, and OpenAI bigtime investor Microsoft, as a non-voting partner — marking the culmination of a messy two-week long odyssey and multiple leadership changes.
Or is it the culmination? In fact, with a new board named and multiple outstanding issues to deal with, another saga is likely just beginning.
What the hell just happened?
If you weren’t following it that closely, here’s the short version: on November 17, 2023, the Friday before Thanksgiving, the previous OpenAI board — comprised at the time of OpenAI CEO and co-founder Sam Altman, OpenAI president Greg Brockman and chief scientist Ilya Sutskever, Quora CEO Adam D’Angelo, technology entrepreneur Tasha McCauley, and Helen Toner of the Georgetown Center for Security and Emerging Technology — had, with the exception of Altman and Brockman, abruptly voted to remove Altman as CEO of OpenAI, stating in a blog post on the company website that Altman was “not consistently candid in his communications with the board,” leading them to lose “confidence in his ability to continue leading OpenAI.”
The move sent shockwaves throughout Silicon Valley and the wider world due to its suddenness, the vagueness of the reasons provided publicly, and the fact that OpenAI is by most accounts the most successful generative AI company and the one most responsible for bringing the technology to the mainstream.
In that subsequent 12-day period of instability and uncertainty, a number of twists and turns occurred, from OpenAI naming two other interim CEOs before finally bringing Altman back, to the vast majority of OpenAI’s staff signing a letter pledging to quit if Altman was not reinstated, to old tweets about controversial sexual fantasies resurfacing and reports of a breakthrough in OpenAI’s quest toward artificial general intelligence (AGI), a machine that performs better than humans at most “economically valuable” tasks. Specifically, a report in Reuters citing anonymous sources suggests OpenAI may have made a major advance towards this goal with a model known as Q* (Q star).
Now, Altman is striking an optimistic tone about the new board and OpenAI writ large in a company blog post, writing he has “never been more excited about the future.”
“I believe our resilience and spirit set us apart in the industry,” he stated, and later, “I am so looking forward to finishing the job of building beneficial AGI with you all—best team in the world, best mission in the world.”
But the new board is still a veritable skeleton crew, halved from just 12 days ago and down from its peak of 10 people in 2021.
Altman’s blog post alludes to the fact that it is likely to grow and contain more “diverse perspectives” — but who else may join remains an open question.
Earlier today, Wired magazine (where my wife is editor-in-chief) reported that prominent and qualified women were not interested in joining the board, repulsed by the apparent “boys club” mentality, and demonstrable chaos of the company. That means finding some said “diverse perspectives” to join up may be a greater challenge than it would otherwise had the saga not occurred.
The only member of the board to survive this latest transition was D’Angelo, whose company Quora and its AI app Poe released a chatbot builder of its own weeks before OpenAI did. Yet Altman took to X on several occasions throughout the saga to praise D’Angelo, stating he spent time with him on Thanksgiving and today to nullify claims D’Angelo was conflicted in serving on the board due to his leadership of a partially rival business.
What happens to OpenAI’s bizarre governance structure?
As VentureBeat senior reporter Sharon Goldman has written, OpenAI is structured in a nontraditional way that appears to have empowered the board to make such an outsized call on Altman in the first place.
Specifically, OpenAI was founded as a nonprofit, but later in 2019 created a wholly owned “capped profit” subsidiary. Now, there are actually multiple subsidiaries that sit under this nonprofit board, only one of which is the actual company that makes AI products and services (another subsidiary is a holding company for OpenAI’s investors and employees and their shares in it — it’s still private for the time being). You can see a chart from OpenAI showing this structure below.
The non-profit board may have had radically different incentives than the “capped” yet for-profit company, exemplified by the board’s focus on candor and safety, while Altman had just days prior to his firing led OpenAI’s first developer conference, DevDay, and launched a series of major new services and initiatives, including the new custom GPT Builder, allowing third-parties to quickly create apps based on ChatGPT using plain English requests, no coding required.
In other words: the nonprofit board seemed to want to move slower, while Altman and his backers including Microsoft and its CEO Satya Nadella, wanted to continue pushing out new products, services, and offerings.
Now that the old board has stepped down and a new one been agreed upon, what happens to the overarching governance structure?
A statement from new board chair Taylor alludes to changes, stating: “We will enhance the governance structure of OpenAI so that all stakeholders – users, customers, employees, partners, and community members – can trust that OpenAI will continue to thrive,” and “As a Board, we are focused on strengthening OpenAI’s corporate governance.”
But exactly how the board changes the governance structure is another big question mark.
What force, if any will Microsoft exert on the new OpenAi board and structure? As a non-voting partner, clearly Microsoft’s role is designed to be limited. But Microsoft CEO Satya Nadella was said to be leading the negotiations between Altman, Brockman, and their supporters and the previous board for Altman to return, was apparently blindsided by and furious about Altman’s initial ouster, and even announced that Altman and Brockman were joining Microsoft to lead a revitalized AI research division before OpenAI ultimately accepted Altman’s boomerang back into CEO.
What happens to Sutskever?
OpenAI chief scientist Ilya Sutskever had led Altman’s firing process, according to a statement by Brockman posted to X, apparently co-authored by Brockman and Altman, recollecting the day of the sudden announcement:
As the statement reads:
- “Last night, Sam got a text from Ilya asking to talk at noon Friday. Sam joined a Google Meet and the whole board, except Greg, was there. Ilya told Sam he was being fired and that the news was going out very soon.
- At 12:19pm, Greg got a text from Ilya asking for a quick call. At 12:23pm, Ilya sent a Google Meet link. Greg was told that he was being removed from the board (but was vital to the company and would retain his role) and that Sam had been fired. Around the same time, OpenAI published a blog post.
- As far as we know, the management team was made aware of this shortly after, other than Mira who found out the night prior.“
Sutskever, an acclaimed AI researcher who has been with OpenAI since its founding year 2015, had sought to defend his actions in an all-hands meeting during a question-and-answer session with confused and upset employees. As reported in subscription tech news outlet The Information last week, Sutskever said Altman’s firing “was the board doing its duty to the mission of the nonprofit, which is to make sure that OpenAI builds AGI that benefits all of humanity.”
Yet a short time later, Sutskever posted on X that he “deeply regret[ted]” his “participation in the board’s actions.”
Now, in Altman’s post, he writes that “I love and respect Ilya, I think he’s a guiding light of the field and a gem of a human being. I harbor zero ill will towards him. While Ilya will no longer serve on the board, we hope to continue our working relationship and are discussing how he can continue his work at OpenAI.”
But how can he continue working alongside Altman and other employees who supported the ousted CEO with the knowledge that their chief scientist plotted against them? Perhaps it is water under the bridge, but it seems like there would likely be unresolved personal tension for some.
Will OpenAI’s old board ever explain to the world why it kicked Altman out?
Among the most striking unanswered questions from the entire Altman vs. OpenAI board leadership fiasco has been the key accusation that started everything: that Altman “was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI.”
The previous OpenAI board still yes to clarify publicly what this means and what Altman is accused of withholding information or lacking candor about. Was it about a breakthrough in AGI that spooked them? About Altman’s rapid pace of development for productizing ChatGPT and its underlying GPT models? Something else?
Short-serving interim CEO Emmett Shear, appointed by the old board members, reportedly sought to receive from them a written reason for firing Altman or he would resign, but they did not provide one.
Yet Shear did post on his X account a long statement including the passage: “The board did not [sic] remove Sam over any specific disagreement on safety, their reasoning was completely different from that. I’m not crazy enough to take this job without board support for commercializing our awesome models.”
Yet just today, U.S. Representative Ro Khanna, a Democrat from California, demanded the OpenAI board testify “before Congress” about “what caused their recent controversy with their CEO and Board of Directors,” stating that “whether it involved their Q* product or some other safety concerns, the public deserves to know.”
If that testimony comes to pass by force of law, say an official Congressional subpoena, it could lead to more information about what really went on behind closed doors at the world’s most successful — and lately, arguably most volatile — generative AI company.
Or, as Altman himself writes in his blog post, “I am sure books are going to be written about this time period.” Perhaps we will need to wait for one of those to find out.
The post OpenAI’s boardroom drama is over… and has just begun appeared first on Venture Beat.