Shulou(Shulou.com)11/24 Report--

Employees revolt, investors sue, allies fight against water. The three outside directors who launched the coup have been cornered.

Author | Xinfu

Editor | Zheng Xuan

Don't underestimate the complexity of things. OpenAI's power struggle doesn't end so easily.

On November 20th, Sam Altman was supposed to be the loser. As the founder and CEO of the high-profile artificial intelligence company, he was abruptly dismissed by the board four days ago and returned to negotiations with the support of employees and investors, but was ignored again. The board directly appointed an outsider as CEO. After that, Altman decided to join Microsoft.

But all OpenAI employees will not allow him to lose the battle-at this critical moment of departure or retention, Altman burst into amazing leadership, and nearly 95 per cent of OpenAI employees signed a collective signature demanding that the board step down and Altman be reinstated-otherwise, they would all resign and join Microsoft. "without employee OpenAI is nothing. "they said.

Employees used this action to make it clear that the core soul of OpenAI is Altman, and without him, the company would fall apart in an instant.

New variables come one after another. The leader of the "contingency" incident, OpenAI board member and chief scientist llya Sutskever has "defected", he appeared at the end of the joint employee letter. On Twitter, he said he regretted the board's action, saying he had no intention of hurting OpenAI and hoped to reunite the company. Altman replied to him with three hearts.

It seems that this power "farce" will come to an end: Altman and Sutskever "let bygones be bygones", and Altman comes back to be CEO. But the remaining OpenAI board members do not seem ready to budge.

Instead of signalling the return of Altman, foreign media reported that they were actively negotiating a merger with rival Anthropic and inviting its co-founder and CEO Dario Amodei to take up the post of CEO of OpenAI-but were rejected.

As of press time, OpenAI employees and board of directors are still in an anxious standoff, and the return of Altman is still a question mark.

01. The board of directors, which has lost its trust, has decided whether Altman can return. Although Sutskever, the leader of the "contingency" incident, has backed out, he cannot bring Altman back. As one of OpenAI's four current board members, Sutskever needs to enlist the support of at least two of the remaining three.

The remaining three members are:

Adam D'Angelo, CEO of American version of Zhihu Quora, founded Poe, a large model product similar to ChatGPT. He joined the OpenAI board in April 2018. "I have always thought that moving towards GM AI requires safety awareness, which is important but neglected, and I am happy to contribute to it. "he said at the time.

Tasha McCauley, a female entrepreneur, co-founded robotics startup Fellow Robots and geospatial technology company GeoSim Systems. She also serves on the advisory board of the International Centre for artificial Intelligence Governance just established in the UK.

Helen Toner, director of strategy at Georgetown University's Center for Security and emerging Technologies, joined OpenAI as a director in September 2021. According to OpenAI's statement at the time, her job was to promote secure and responsible technology deployment.

Four days ago, Sutskever joined forces with the three directors to expel Altman and Greg Brockman (the former OpenAI president and board member, who was also ousted as a director in this incident, and then he announced his resignation). But now, it doesn't seem so easy to get the support of these three and get Altman and Brockman back.

We analyzed in yesterday's article why OpenAI should let outsiders serve as the core board member of the company. The core is that in order to maintain the position of the non-profit organization, OpenAI stipulates that the profit subject of the company shall be managed by the board of directors of the non-profit body of the company. In the board of directors, only members who do not hold non-profit shares have the right to vote.

This time, the board announced the removal of Altman on the grounds that he could not always communicate honestly with the board, hindered the board's performance of its duties, and that the board no longer believed in his ability to continue to lead OpenAI. The board believes that it is exercising its responsibility as a non-profit organization, "that is, to ensure that OpenAI builds AGI for the benefit of mankind as a whole".

However, due to the various practices of the board of directors, its motivation and mission have been questioned by employees as never before.

According to foreign media reports, the senior management of OpenAI repeatedly asked the board of directors to explain in detail the reasons for their dismissal of Altman and provide corresponding evidence, but the board of directors did not give an explanation. The employee also confirmed this in a joint letter, "although many people have asked you to provide specific factual allegations, you have never provided any written evidence." "they wrote.

At about the same time that employees released the joint letter, Microsoft CEO Satyanadra (Satya Nadella) told Bloomberg television that he knew nothing about the board's objections to Altman and that he had never heard of any problems with Altman from anyone on OpenAI's board.

"so I still have confidence in Sam and its leadership, which is why we welcome him to Microsoft. "he said.

It is widely speculated that the root of Altman's disagreement with the board is that his aggressive technology, products and commercialization strategy clash with the board's security concerns about technology and the non-profit mission.

But more details seem to be shattering that speculation. According to foreign media reports, an OpenAI board minutes said that the decision to expel Altman "has nothing to do with product safety, the speed of development or the financial situation of OpenAI." "

The OpenAI interim CEO Emmett Shear appointed by the board (he was a co-founder of the video streaming site Twitch) posted that he had learned about the reason why Altman was ousted, not because the two sides claimed to have specific differences on AI security issues. "that's not the reason at all. I'm not crazy enough to take the job without the board's approval of a commercial model. "

Vague reasons and evidence, and so tough and conflicting in specific actions-OpenAI employees were haunted by the fact that after the announcement of the removal of Altman, the board of directors did not accept proper communication with OpenAI management, but directly replaced the temporary CEO Mira Murati (OpenAI CTO) and directly introduced an outsider Shear as CEO--, which completely lost the trust of the employees.

"you [the board] are incapable of performing your duties and are full of malice in the negotiations. Your behavior has shown that you are unable to control OpenAI. We cannot work with people who lack ability, judgment and care for our mission and employees. "the employee said in the open letter.

However, in the face of employee complaints, the board of directors secretly negotiated the acquisition with the company's competitor Anthropic, and it can be expected that the current mistrust and conflict between employees and the board of directors may be approaching a climax.

That's where OpenAI is now: the board, which has real power and decides the fate of the company, first expels the soul of the company, and then completely loses the trust of its employees in communication and action, which is doing its best to overthrow it.

02, Altman and Microsoft wait for the results, competitors desperately poach customers at a time of fierce confrontation between OpenAI's board of directors and internal employees, Altman, the core figure of this incident can only wait and see-employees win, he is expected to go back to the OpenAI; board to win, and he can sit on all the original teams and open up his own new mountain at Microsoft.

But both he and Microsoft, of course, prefer the first result, which avoids the cost of moving resources and wasting resources. According to The Verge, Altman said he would still be willing to return to OpenAI if board members give in.

Nadella made it clear in a Bloomberg television interview that Altman is likely to return to OpenAI. He added that Microsoft will benefit whether Altman returns to OpenAI or works at Microsoft.

According to foreign media reports, OpenAI investors, who saw the invested companies in a major crisis, are seeking to sue the OpenAI board of directors.

Vinod Khosla, OpenAI's first venture capitalist and founder of Khosla Ventures, wrote that the wrong board could hurt companies and that the best companies were "companies whose vision is led and implemented by their founding entrepreneurs".

Joshua Kushner, another investor in OpenAI and the founder of Thrive Capital, wrote an article in support of OpenAI employees on X (formerly Twitter). Founders should have the right to run their own business. "he wrote.

Within OpenAI, the new interim CEO Shear seems interested in making some communication and changes between the conflict-prone board and employees.

In a tweet after taking office, he said he saw "very poor handling of the process and communication surrounding the dismissal of Sam", which "seriously damaged our trust [between the board and employees]. "

He proposed a "three-point plan" for the company over the next 30 days: first, hire independent investigators to learn more about the process so far and generate a complete report; second, talk to employees, partners, investors and customers as much as possible, keep records, and share key points. Third, if necessary, strongly promote major changes in corporate governance.

But according to foreign media reports, the temporary CEO was ignored by internal staff.

According to foreign media reports, at a time when OpenAI is engaged in fierce infighting, its competitors such as Amazon and Anthropic are actively discussing how to pry away its big customers. There is no need to even pry, and at least 100 OpenAI customers are approaching competitors because of concerns about fierce fighting within the company.

This is OpenAI's most critical battle-- as the world's top and most believing artificial intelligence company, it faces the management chaos caused by the corporate governance structure for the first time, and its first collision with the real and complex business world-- and whether it can survive this difficulty smoothly determines not only the fate of the company, but also where the future AI business landscape will go.

This article is from the official account of Wechat: geek Park (ID:geekpark), author: Xinfu

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.