Wall Street is growing increasingly impatient with Meta CEO Mark Zuckerberg’s secretive push into advanced artificial intelligence, as he pours billions into a mysterious new “superintelligence lab” while offering enormous incentives to lure top AI talent from rivals. As reported by the Financial Times, Zuckerberg is orchestrating an aggressive hiring blitz — complete with nine-figure signing bonuses — to build a start-up-like unit within Meta, isolated from the broader company and reporting directly to him. Despite the spending surge, Meta has provided few details to investors or even internal staff about the lab’s mission, fueling concerns over direction, returns, and accountability.
The lab, headquartered in a separate facility at Meta’s Menlo Park campus, is led by Alexandr Wang, former CEO of Scale AI, and Nat Friedman, ex-GitHub chief. Its existence underscores Zuckerberg’s ambition to leapfrog AI rivals and reclaim dominance after Meta’s Llama models underperformed compared to offerings from OpenAI and others. This internal frustration has led Zuckerberg to empower a handpicked team of roughly 50 elite researchers with the task of developing superintelligence — AI systems more advanced than human cognition — while simultaneously guiding Meta’s overall AI efforts.
Despite the opaque strategy, investor enthusiasm hasn’t yet cooled; Meta’s stock is up 20% this year. However, with the company expected to report its slowest profit growth in two years and its capital expenditure forecast already raised to as much as $72bn, analysts are pressing for clarity. Brent Thill of Jefferies cautioned that while the hiring spree indicates urgency, it could “pressure the bottom line” in the near term. BNP Paribas estimates the initiative may add $1.5bn to $3.5bn annually in research and development costs.
The stealthy nature of the lab — which has even been compared by one investor to a “Manhattan Project” — has stirred unease inside Meta, where existing AI researchers fear marginalization. One staffer told the Financial Times that many feel Zuckerberg is attempting to “replace” them, creating friction as the company struggles to find a coherent, resonant mission in AI. Meta’s longstanding AI leader Yann LeCun remains at the company but now reports to Wang, raising questions about how responsibilities are being reallocated behind closed doors.
Meta’s latest AI model, Llama 4, launched to tepid reception, drawing criticism from developers and leading Zuckerberg to delay the rollout of its larger Behemoth model until it meets his higher expectations. Frustrated with slow internal progress, he has reportedly allowed Meta teams to experiment with external models, further signaling dissatisfaction with current capabilities. According to insiders, this shift toward speed over legacy systems reflects Zuckerberg’s urgent desire to catch up with competitors like OpenAI, Anthropic, and Google, all of which have been pouring billions into the AI race.
In a rare public comment, Zuckerberg defended the strategy in an interview with The Information, saying Meta’s vast cash-generating business model gives it the freedom to make long-term bets. He stressed that such initiatives should be kept small and tightly knit to foster innovation, stating, “You actually kind of want the smallest group of people who can fit the whole thing in their head.”
The talent war sparked by Meta’s AI offensive has rattled Silicon Valley. The company successfully poached high-profile names including Shengjia Zhao, a co-creator of ChatGPT, now appointed chief AI scientist of the new lab. Meta also attempted — unsuccessfully — to acquire start-ups founded by ex-OpenAI executives, though it did secure Daniel Gross, co-founder of Safe Superintelligence. OpenAI CEO Sam Altman took a swipe at Meta’s tactics, deriding them as “mercenaries” lacking purpose beyond profit.
Meta’s prior big bet — the metaverse — proved a costly disappointment, eroding investor confidence and tanking its share price in 2022. Now, Zuckerberg’s pivot to AI is being watched closely by both markets and employees. Analysts at MoffettNathanson note that while Meta’s AI strategy is “more cohesive” than it was a year ago, it still feels adrift, without a clear product roadmap or business model that investors can rally behind.
As Meta prepares to release its quarterly earnings, stakeholders await more than numbers — they want visibility into a vision. Without it, even Zuckerberg’s boldest AI gambit may struggle to inspire the confidence required to sustain the enormous investment it demands.
