Dismissal, uprising, reinstatement: the OpenAI Sam Altman crisis narrative

The abrupt firing of the ChatGPT company’s CEO sent shockwaves through Silicon Valley, only to see him reinstated days later

Nine days ago, Sam Altman, the CEO of OpenAI, hinted at a significant development in the world of artificial intelligence during a speech in San Francisco. He expressed the professional honor of being involved in pushing the frontier of discovery forward. However, the following day brought a surprising turn of events as Altman was fired from his position, leading to speculation about the future of AI at OpenAI.

On November 17, OpenAI’s board announced the dismissal of Sam Altman, citing a lack of “consistently candid” communication. Altman received a text notifying him of a discussion with fellow board member Ilya Sutskever, OpenAI’s chief scientist, scheduled for noon the following day via Google Meet. During the call, Sutskever informed Altman of his termination, as detailed in a post on X (formerly Twitter) by Greg Brockman, OpenAI’s chair and president.

Brockman, who was not present during Altman’s dismissal call, received a summons minutes later for a Google Meet discussion where he learned about his removal from the board. Although offered the option to remain with the company, Brockman chose to resign. OpenAI then named Mira Murati, the chief technology officer, as interim CEO.

The news of Altman’s termination created shockwaves in Silicon Valley, prompting immediate calls from OpenAI investors for his reinstatement. OpenAI, initially established as a non-profit with a board overseeing its commercial subsidiary, counted Microsoft as its largest investor. Satya Nadella, Microsoft’s CEO, revealed that his company had not been consulted before the announcement.

On X, Altman expressed his affection for OpenAI on the same Friday when his termination was announced, but it quickly became evident that the saga was far from concluded. The Information, a tech news platform, cited a memo from Jason Kwon, the chief strategy officer, informing the staff on Saturday about ongoing efforts to reinstate Altman and other departed colleagues, including the three senior researchers Jakub Pachocki, Aleksander Mądry, and Szymon Sidor.

“We are continuing efforts for a resolution, and our optimism persists,” Kwon stated. “By resolution, we mean bringing back Sam, Greg, Jakub, Szymon, Aleksander, and other colleagues.” Sunday introduced more unexpected developments. Altman shared an image of himself within OpenAI’s San Francisco headquarters, holding a guest pass, accompanied by the message, “first and last time I ever wear one of these.

The discussions with the remaining four members of OpenAI’s board proved unsuccessful, leading to the appointment of Emmett Shear, the former CEO of the video game streaming platform Twitch, as the company’s third CEO in three days. Shortly thereafter, Nadella revealed that Microsoft had recruited Altman and Brockman to spearhead a “new advanced AI research team.” However, in subsequent media interviews on Monday, Nadella expressed openness to Altman’s potential continuation at OpenAI. An analyst highlighted that due to Microsoft’s investment, Altman would effectively hold a leadership role at OpenAI.

By Monday, the underlying discontent among staff regarding Altman’s removal erupted into open rebellion. The majority of OpenAI’s 770 employees, including Murati, joined in signing a letter addressed to the board, threatening a collective resignation unless Altman and Brockman were reinstated – a condition that would necessitate the subsequent resignation of the board.

The letter, signed by numerous staff members, including surprising support from board member Sutskever, stated, “We are unable to work for or with people that lack competence, judgment, and care for our mission and employees.”

The three other board members, namely Adam D’Angelo, the co-founder of the question-and-answer website Quora; Tasha McCauley, a tech entrepreneur; and Helen Toner, a director and AI safety expert at Georgetown University’s Center for Security and Emerging Technology, were also signatories. However, indications emerged that Altman was moving closer to a potential reinstatement. OpenAI’s vice-president of global affairs, Anna Makanju, informed staff that “intense discussions” were underway to resolve the crisis. Bloomberg reported that Murati communicated to staff on the Slack platform, stating, “still working on it.

Then, a deal was brokered. OpenAI, on Tuesday night in California, declared that an agreement in principle had been reached for Altman’s return, accompanied by a revamped “initial” board chaired by Bret Taylor, the former co-chief executive of software company Salesforce, along with Larry Summers, the former US Treasury secretary, and D’Angelo. Altman has not, as of now, regained his position on the board.

OpenAI is reportedly set to conduct an independent investigation into the circumstances surrounding Altman’s departure, initially announced by Shear. Shear had previously denied that Altman’s dismissal was linked to safety concerns.

However, following Altman’s return, Reuters and the Information reported that some OpenAI researchers had expressed concerns about the capabilities of a new AI model being developed by the company before the CEO’s termination. According to the Information, the “Q*” model – pronounced “Q-star” – demonstrated the ability to solve previously unseen math problems, potentially indicating a breakthrough in the technology.

Established in 2015, OpenAI aimed to develop “safe and beneficial artificial general intelligence [AGI] for the benefit of humanity.” AGI denotes an AI capable of performing various tasks at, or surpassing, human intelligence levels.

Concerns exist among experts that the advent of AGI could theoretically result in a system evading human control, posing a threat to humanity. While OpenAI’s board has undergone changes, its mission remains consistent.

Altman, who reportedly sought a new board as a condition for his reinstatement, hinted on Thanksgiving in the US (Thursday) that he was actively working to mend any rifts with D’Angelo.