In a bold and unexpected shift, Meta CEO Mark Zuckerberg has stepped into the driver’s seat of a secretive and ambitious project: building a superintelligent AI. According to a detailed Bloomberg report, Zuckerberg is now personally spearheading efforts to develop a system that could potentially surpass human intelligence.
Meta has long integrated artificial intelligence across its product suite, from Facebook and Instagram to WhatsApp, chatbot assistants, and even Ray-Ban smart glasses. Yet, despite these implementations, the company has struggled to keep up with the momentum of rivals like OpenAI, Microsoft, and Google.
Sources close to the matter suggest that Meta’s latest large language model, Llama 4, failed to meet Zuckerberg’s expectations, primarily due to sluggish development. That frustration has reportedly prompted the CEO to form a compact yet elite internal team of around 50 engineers and researchers, stationed near his own office in Menlo Park. This team will work under his direct oversight, marking a new, hands-on chapter in Meta’s AI journey.
Zuckerberg’s vision is clear but immensely ambitious: to create a system that achieves “superintelligence” AI that outperforms human cognition. But before this goal can be realized, Meta must first build artificial general intelligence (AGI), a system capable of matching human reasoning and problem-solving skills across a wide range of tasks.
This is no simple feat. AGI is still a largely theoretical concept, and while some researchers believe it could be achieved in a matter of years, others caution that we may be far from cracking its code. Even so, Zuckerberg appears undeterred, seeing AI not just as a field of innovation, but as a make-or-break factor for Meta’s long-term survival.
To support this monumental effort, Meta plans to leverage its powerful advertising engine for funding. Part of its strategy reportedly involves a potential multibillion-dollar investment in Scale AI, the data-labeling company led by 28-year-old Alexandr Wang. The New York Times recently confirmed Wang’s involvement in the superintelligence initiative, adding another layer of intrigue to the project.
Zuckerberg’s fixation on AI intensified after OpenAI’s ChatGPT reshaped the public and tech industry’s understanding of AI potential. Meta’s move to open-source Llama was meant to mirror Android’s open ecosystem and attract developers globally. However, despite those efforts, the results have not yet matched the disruptive force of ChatGPT or Google’s Gemini models.
Now, the stakes are even higher. Meta isn’t just contending with OpenAI or Google, it’s up against a growing list of players, including Anthropic, Elon Musk’s xAI, and even Apple, which recently unveiled its delayed entrance into the AI space.
For Meta, the AI arms race isn’t just about technological prestige it’s about survival. Just as Google fears AI could replace traditional search, and Apple is preparing for a future where AI might render apps obsolete, Meta sees AI as both a threat to its current business model and an opportunity to reshape its role in the tech landscape.
OpenAI may currently dominate the narrative, but Meta is banking on Llama’s open-source appeal to attract a vibrant AI community. If it succeeds, Llama could serve as the underlying infrastructure for a global AI ecosystem.
Still, the finish line remains unclear. As Zuckerberg pushes deeper into uncharted AI territory, experts remain divided on just how close we are to unlocking AGI or superintelligence. And even if those capabilities become technically possible, their societal, ethical, and economic consequences are far from settled.
