Now that the incredible episode of layoffs at OpenAI has subsided, tongues are loosening as to the reasons which pushed the management committee to part ways with Sam Altman.
The month of November was not easy for the parent company of ChatGPT. After the surprise dismissal of its CEO, threats of massive migration among competitors and the return to favor of the co-founder, the company is just starting to get back into working order and reconstitute its management committee. New revelations allow us to see more clearly in this troubled episode.
A philosophical tear at OpenAI
Charles Duhigg, journalist for the New Yorker magazine, recently told in an article why Sam Altman was let go from OpenAI in the snap of his fingers. While the official reason mentioned so far was “ lack of sincere communication » from the CEO, the full story is more strategic and political than that.
A few weeks before Altman’s firing, Helen Toner, an OpenAI board member and director of strategy at Georgetown’s Center for Security and Emerging Technologies, published an article accusing OpenAI “to maintain a frenzy around AI» despite risks that are still poorly identified. Supporting a philosophical line called “effective altruism” (or “effective altruism” in French), the researcher supports the idea that each evolution of AI must be weighed and measured against the risks it could create in society. .
Annoyed by the publication of the article and by the line defended by Helen Toner, the CEO of the company then began to pit the members of the management committee against each other to obtain the ejection of the researcher. Once Altman’s ruse was exposed, these same members would have met (in secret) to discuss the CEO’s departure because of the methods he had demonstrated.
Altman back in charge, stronger than ever
“This sort of thing has been commonplace for years», Explained an anonymous source to Charles Duhigg. According to the journalist’s account, Altman indeed had the reputation of being a particularly “rogue», hence the need to meet in secret to prevent the CEO from undermining the authority of the management committee. Another source who spoke to Charles Duhigg, however, defended Sam Altman, calling this whole internal mess “clumsy operation» rather than an attempt at manipulation.
“The mission is multifaceted, it is to ensure that AI benefits all of humanity, but no one can achieve this without holding the CEO accountable» summarized a source within OpenAI.
The rest of the story is known: after many twists and turns, Sam Altman took over the helm of OpenAI and Helen Toner was dismissed. Now sitting on the management committee are Bret Taylor, Larry Summers and Adam D’Angelo, three men who will have to stand up to the charismatic CEO.
Source : New Yorker
3