In a shocking move and pivotal moment for AI last Friday, Sam Altman, the co-founder and CEO of OpenAI, was fired by its board in a vague and cold letter saying he wasn’t “consistently candid in his communications with the board.”

Three CEOs in 72 hours: The board briefly chose Chief Technology Officer Mira Murati as interim CEO, but turns out, Murati was team Altman — and had tried to rehire him. By late Sunday evening, OpenAI’s board had changed their mind — naming former CEO of Twitch Emmett Shear, as CEO.

After Altman’s ousting, COO Brad Lightcap sent an internal memo clarifying that the decision wasn’t due to “malfeasance or anything related to our financial, business, safety, or security/privacy practices.”

It’s the opposite: The company is becoming too successful. Last week, OpenAI had to pause signups to its premium product due to overcapacity. Within eight years, Sam Altman had transformed OpenAI into a company potentially worth $90B, three times what it was valued at just a year ago.

• Satya Nadella, CEO of OpenAI’s largest investor, Microsoft (NASDAQ:MSFT), was “blindsided” by the decision and pledged to support Altman in whatever decision he makes.
• Co-founder Greg Brockman and three senior researchers quit in protest — while OpenAI investors pushed to reinstate Altman and replace the board.

Some even reportedly devalued their OpenAI investment to zero over the weekend, pressuring the company to bring Altman back. OpenAI investors (+Altman) had considerable leverage, considering OpenAI had only received a portion of Microsoft’s $10B investment — with a part of it taking the form of Microsoft’s cloud computing services. But even that wasn’t enough.

Breaking: Early Monday morning, Nadella announced in a LinkedIn post that Altman and Brockman had joined Microsoft to “lead a new advanced AI research team.” Turns out, Microsoft was the one to come out on top here.

AI taking over doesn’t seem too farfetched…

Internally, OpenAI has become increasingly divided over whether to push for the business’ expansion — or prioritize safety. Co-founder Ilya Sutskever, a vocal advocate for AI safety, established a “Superalignment” team within OpenAI to ensure responsible AI use.

• Sutskever sided with board members, which consisted of scientists worried about the rapid development of AI’s impact on safety.
• Defending the board’s decision, Sutskever said, “This 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” (The Information).

Let’s take a moment of silence for AI safety — as Meta also disbanded its Responsible AI team last week.