Meta’s aggressive hiring spree in artificial intelligence is shaking the company’s stable leadership core, contributing to the Meta AI shake-up. It is leaving longtime insiders unsettled. Meanwhile, newcomers reorient the social media leader’s direction. Reports suggest CEO Mark Zuckerberg is pushing harder than ever to build what he calls “personal superintelligence”. However, the process is messy, and not everyone is staying on board.
What’s Happening & Why This Matters
Within days of joining Meta, Shengjia Zhao, co-creator of OpenAI’s ChatGPT, nearly walked out the door to return to his former employer. Instead, he was named chief AI scientist and given control of Meta’s superintelligence program, highlighting the Meta AI shake-up. The appointment notes the value of Zuckerberg’s latest strategy. It also highlights the volatility of poaching top talent with nine-figure packages and access to vast computing resources.
The shake-up is the most dramatic reorganisation of Meta’s senior ranks in two decades. Longtime allies like Chief Product Officer Chris Cox have been pushed to the sidelines. Rising stars such as Zhao, Alexandr Wang (formerly of Scale AI), and Nat Friedman (ex-GitHub) are taking centre stage in this Meta AI shake-up. These hires now oversee the newly formed Meta Superintelligence Lab (MSL). It has been reorganised yet again into four divisions. This marks the fourth major restructuring in just six months.
Yet, despite the glamour of high-profile signings, attrition is already underway. Some recruits, including Ethan Knight and Avi Verma, have already departed or never showed up. Others, like Rishabh Agarwal, joined only to announce swift exits. Meanwhile, respected veterans such as Chaya Nayak and Loredana Crisan, who each spent nearly a decade at the company, are also leaving, illustrating part of the ongoing Meta AI shake-up.
Meta defends the churn, describing it as normal for a company of its size. It insists it remains focused on delivering superintelligence. Still, insiders describe an organisation riddled with bureaucracy and resource bottlenecks. There is growing frustration among staff promised powerful infrastructure that has yet to materialise.
An Internal Culture Clash
At the heart of the tension is Zuckerberg’s management style. While some describe him as deeply invested, others criticise him for micromanaging and pushing aggressive timelines for breakthroughs in AI. Reports indicate he is pressuring Wang. Wang oversees the most secretive department — fittingly labelled “TBD” — and is pushed to deliver results faster than some researchers believe possible.
Adding to the pressure, the company is said to be quietly shelving public release plans for its flagship Llama Behemoth model after disappointing results. Instead, resources are being redirected. The focus is now on the development of new, more advanced systems.
Notably, Yann LeCun, Meta’s longtime chief AI scientist, remains in place but now reports to Wang. This is a clear indicator of new power dynamics. Others, such as Ahmad Al-Dahle, who once spearheaded generative AI efforts, have been left without clear leadership roles.
Despite the turbulence, some in Silicon Valley view the hiring of Zhao and Friedman as smart bets that could pay off. With Meta investing billions into AI infrastructure, the company still has the resources and reach to challenge rivals like Google DeepMind, OpenAI, and Anthropic.
TF Summary: What’s Next
Meta’s AI gambit is equal parts ambition and uncertainty. The aggressive hires and reorganisations show Zuckerberg is determined to leapfrog competitors. But constant reorgs and high-profile exits risk slowing momentum. If the new leadership team stabilises, Meta could make real progress toward its superintelligence ambitions. If not, the reshuffle could be a cautionary tale in how not to build an AI future amidst a Meta AI shake-up.
— Text-to-Speech (TTS) provided by gspeech