Meta is facing a crisis as a series of AI experts, including former OpenAI and Google employees, have left Mark Zuckerbergs AI dream team. Can Meta still maintain its advantage in the race for artificial superintelligence?
Meta once made waves in the AI technology world when CEO Mark Zuckerberg personally 'headhunted', inviting top experts from OpenAI, Google and Apple with contracts worth millions of dollars. However, just a few days after the news that the company was forced to temporarily suspend recruiting more AI personnel due to pressure from investors, Meta is facing another problem: retaining new faces for the 'Superintelligence Lab' is not easy.
According to Ars Technica, Shengjia Zhao – co-founder of ChatGPT – was about to leave Meta just days after joining. He even signed his resignation papers before suddenly changing his mind, accepting the title of 'Chief AI Scientist' and staying. In another case, scientist Ethan Knight left the company just weeks after joining. Former OpenAI expert Avi Verma completed his onboarding but did not show up on his first day. Rishabh Agarwal, who joined in April, was initially attracted to Zuckerberg's vision but then decided to leave because he wanted to 'take a different kind of risk'.
Not only new faces, a series of veteran staff also announced their departure from Meta. Among them are Chaya Nayak and Loredana Crisan - two key members of GenAI with nearly 20 years of combined experience.
From an outside perspective, the situation appeared chaotic, but Meta was quick to dismiss it as 'hype':
We understand there is a lot of attention to detail in our AI efforts, but the important thing is that we are focused on building personal superintelligence. It is normal for some people to leave at any large organization. Most of them have been with the company for many years, and we wish them all the best.
Zuckerberg's ambitions for 'personal superintelligence' are being realized through the establishment of the Meta Superintelligence Lab, which consists of four teams: a 'TBD Lab' focused on superintelligence research from new employees, an AI product development team, an infrastructure team, and a third team in charge of long-term ideas. This force is currently led by Zhao, technology entrepreneur Alexandr Wang, and former GitHub CEO Nat Friedman.
However, the emergence of these 'new stars' has also created a lot of internal conflicts. Yann LeCun - one of the veteran AI scientists - is now "under the direction" of Wang. Meanwhile, Ahmad Al-Dahle, who played an important role in the development of Llama and many GenAI projects, is no longer leading any team. Notably, Wang reports directly to Zuckerberg, bypassing Chief Product Officer Chris Cox in the management chain.
Some sources also said that Zuckerberg and Wang were at odds over the timeline for achieving superintelligence, but Meta denied it, calling it 'baseless dramatic stories created by hypemen'.
Whether Meta can realize its ambition of building personal superintelligence remains a big question mark. But clearly, Mark Zuckerberg's AI 'house' is facing significant shakeups.