2023-11-22 18:41:00
A scenario worthy of Hollywood. This early Wednesday morning, Sam Altman regained his position as CEO of OpenAI and thus put an end to an imbroglio never seen before in the history of Silicon Valley. For five days, the company that created ChatGPT, the new star of American tech, changed CEOs three times to ultimately… return to the status quo.
If the management of OpenAI has not changed, the incredible saga will still leave its mark on the governance of OpenAI. The board of directors that initiated Altman’s removal is no more, and the new board put in place this Wednesday is intended only to ensure the transition before the appointment of an expanded board. With one mission: that such a fiasco does not happen again, while the company aims to create ever more powerful, and therefore dangerous, intelligences.
November 17, coup d’état at the head of OpenAI
Friday evening, to everyone’s surprise, OpenAI published a press release to announce the dismissal of its manager Sam Altman. Co-founder of the organization, he had held his position since its creation in 2015. The shocking news: the entrepreneur had established himself as the face of the ChatGPT revolution and even the incarnation of the new wave of artificial intelligence, which attracts so much desire. Better still, he had led the company towards an expected valuation of $90 billion, multiplied by three in less than a year.
At the origin of this coup are four of the five members of the OpenAI board of directors. Ilya Sustkever, the company’s most renowned scientist, considered the main architect of the success of OpenAI’s research, becomes the spokesperson for the rebels. It was he who announced his dismissal to Sam Altman, by videoconference, a few minutes before the publication of the press release. It was also he who warned Greg Brockman, another co-founder of the company, of his dismissal as chairman of the board of directors of which he was the fifth member. In the process, the latter resigned, despite the fact that OpenAI wanted to keep its expertise within its confines.
The brutality of the decision triggers chaos at OpenAI and indeed the entire artificial intelligence ecosystem. The confusion reigns all the more as the council remains evasive on the reasons for its decision. He criticizes Altman for “ not always being sincere in your communications with the board ”, but refuses to give the details, whether to employees or investors. Moreover, the latter are fuming about not having been warned in advance, like Microsoft, which invested more than 13 billion dollars in OpenAI. Mira Murati, the CTO of the company renowned for its management of the ChatGPT project, is named interim CEO.
One year after ChatGPT, the dawn of a new era
A decision driven by a war of ideas?
A question emerges: how could a board of directors fire a CEO whose actions benefited investors so much? The answer lies in the structure of OpenAI: the board has the governance of the original, non-profit structure. Its mission is not to optimize investors’ gains, but simply to ensure that the AI developed by OpenAI benefits humanity, and does not stray into dangerous territory. The problem is that this council has control over the commercial structure of OpenAI, created in 2019 in the wake of Microsoft’s investment.
Could the coup therefore have had an ideological origin? Within OpenAI, Sutskever embodies the organization’s original mission: to create artificial intelligence capable of reasoning similar to or even superior to that of humans, but in a controlled manner. He also leads the team of “ superalignment », responsible for ensuring that AI systems are not dangerous for humans. With the race for product and performance triggered by ChatGPT, the scientist has several times expressed his concerns about the lack of precautions in the deployment of AI. And he would have confronted Sam Altman on this subject, great architect of the new services presented by OpenAI at the beginning of the month, during his first annual conference. Reported by the American press, this conflict of ideas was however never publicly confirmed by the board of directors, which stuck to its original position.
November 19, second slap for Sam Altman
On Sunday, with the support of investors and the support of employees, Sam Altman went to his former premises to apply for his own succession and resolve the disagreement that led to his dismissal. But the council of rebels decides to inflict a second slap in the face, and to appoint Emmet Shear, former head of Twitch, as CEO. An anti-Sam Altman, supporter of very careful development of AI. As soon as his appointment was announced, the latter explained that the speed of development of AI should be divided by at least five, for security reasons. Once again, the board makes its decision without consulting investors…
After this new affront, Sam Altman, accompanied by a handful of OpenAI executives including Greg Brockman, announced that he was joining Microsoft. The tech giant, which enjoys a privileged partnership with OpenAI, says it is ready to create a tailor-made research division to accommodate the exiled leader and all the employees who would like to follow him. The soap opera then reaches the peak of drama.
November 22: Sam Altman more powerful than ever
On X, employees coordinate their communication operation, in support of their former CEO. They sing in chorus “ OpenAI is nothing without its people » [OpenAI n’est rien sans ses employés, ndlr], and react with the heart emoji to Sam Altman’s messages. As of Sunday evening, several hundred employees, including Mira Murati and Ilya Sutskever, signed a letter to warn the board of directors that without return from Sam Altman, they would resign and join their beloved leader at Microsoft. With more than 500 signatures from 770 employees upon its publication, the letter puts the knife to the throat of council members. Especially since Monday evening, it exceeded 700 signatures. If the council does not bend, it risks governing an empty shell. This is the moment that Ilya Sutskever, considered the leader of the revolt, chose to publicly apologize on X (formerly Twitter). He regrets his participation in the coup, and explains that he will do everything in his power to reunify the company. The icing on the cake was that Microsoft CEO Satya Nadella declared on television Monday evening that he would leave the door open for Sam Altman to return to his former role, but that OpenAI would need new governance.
It was then only a matter of time before Sam Altman was back in office, with unanimous confidence from his employees. The company has chosen a new board of directors who is responsible for ensuring the transition to a new permanent board of directors, OpenAI opens a new page. And its emblematic leader now has confirmation that he is indestructible, while ChatGPT celebrates its one year anniversary next week.
One year of ChatGPT: prove to me that you are still human!
1700967549
#reasons #triumphant #return #Sam #Altman #star