AI in practice

OpenAI crumbles in just three days, leaving more than 700 employees wondering what went wrong

Matthias Bastian
Sam Altman selfie

Sam Altman via Twitter.com

Update
  • Added statement from Satya Nadella about OpenAI staff joining Microsoft

Update from November 21, 2023:

CEO Satya Nadella commented on OpenAI's future plans in interviews with CNBC and Bloomberg TV. When asked if OpenAI employees would move to Microsoft, Nadella replied that this is a decision for OpenAI management and employees. Microsoft is open to both possibilities. The OpenAI employees are in the process of being brought on board - Nadella was not more specific about what that means.

Regarding Microsoft's participation in the OpenAI board, Nadella said it was clear that something had to change in terms of governance. He emphasized that surprises are bad and that Microsoft definitely wants some changes in governance. Nadella left open who the CEO of OpenAI will be, citing the decision of the OpenAI board.

Original article from November 20, 2023:

It only took three days for OpenAI to go from the hottest AI company in the world to total collapse.

About 650 of OpenAI's 770 employees are threatening to quit their jobs after Sam Altman's sudden ouster last Friday, writes OpenAI employee Lilian Weng on Twitter.com. (Update: Now 730 employees, according to Will Depue.)

In an open letter, the employees vent their frustration with the OpenAI board, which, according to the letter, has yet to provide a clear explanation for Altman's dismissal.

While the OpenAI management team has been working with the board, no solution has been found.

Instead, the board has even admitted that destroying OpenAI could be in line with OpenAI's nonprofit mission to develop AGI for the benefit of humanity.

The board has shown a lack of competence, judgment, and care for OpenAI's mission and employees, the letter states. The staff is threatening to move to Microsoft's new AI unit, led by former OpenAI CEO Sam Altman if the board does not resign.

Image: OpenAI letter, via Kara Swisher on Twitter.com

The letter is signed by several senior executives, including former CTO Mira Murati, who served as interim CEO for less than two days after Sam Altman's departure, COO Brad Lightcap, and co-founder John Schulman.

Oddly enough, OpenAI researcher Ilya Sutskever, who has been blamed for Altman's downfall, also signed the letter. On Twitter.com, Sutskever expressed regret for participating in the board revolt. "I will do everything I can to reunite the company," Sutskever writes.

Image: Ilya Sutskever via Twitter.com

There is still a chance that Altman and Brockman will return to OpenAI

But it gets even better: Ex-CEO Sam Altman and Greg Brockman, co-founder of OpenAI, are apparently still willing to return to OpenAI if the remaining board members step down.

According to The Verge, Altman's move to Microsoft has not yet been finalized. The possible departure of OpenAI employees, including board member and chief scientist Ilya Sutskever, increases the pressure on the remaining board members to step down.

Altman, Brockman and OpenAI's investors were looking for a suitable exit for the board, while Microsoft's hiring announcement was considered something of a holding pattern, according to the report.

"We are all going to be working together some way or another, and I'm so excited. One team, one mission," Altman wrote on Twitter.

He added that the close partnership between Microsoft and OpenAI means that he will continue to be involved in the development of OpenAI. All services will continue.

Meanwhile, The Information reports that more than 100 OpenAI customers have inquired about competitor Anthropic, which recently raised billions from Amazon and Google. Other customers are considering switching to Google Cloud, Cohere, or Microsoft Azure, which offer copies of the OpenAI models.

Altman's firing allegedly has nothing to do with AI safety

OpenAI's alleged new CEO, Emmett Shear, also weighs in, presenting his 30-day plan, which includes an investigation into the events of the past few days and a reorganization of management.

Shear also says that Altman's firing had nothing to do with a "specific dispute" over AI safety, as had been suspected and reported by internal sources. The reason was something else entirely, Shear said, without revealing it.

"I'm not crazy enough to take this job without board support for commercializing our awesome models," Shear wrote.