Meta is quietly building what could become the most ambitious artificial intelligence project in Silicon Valley history. The social media giant is launching a new research facility focused on “superintelligence“—AI systems that could surpass human cognitive abilities.
The stakes couldn’t be higher. After spending $46 billion on the metaverse with limited results, CEO Mark Zuckerberg is placing an even bigger bet on machine learning and advanced AI development.
The talent war reaches fever pitch
Meta has recruited Alexandr Wang, the 28-year-old founder of Scale AI, to lead this groundbreaking initiative. The company invested $14.3 billion in Wang’s startup and brought his entire leadership team aboard.
But that’s just the beginning. Meta is offering compensation packages worth $100 million to $200 million to top researchers from OpenAI, Google, and other AI companies. These unprecedented salary offers are reshaping the entire industry’s compensation standards.
“It is a very hot market,” Zuckerberg acknowledged during a recent interview. “There’s a small number of researchers, who are the best, who are in demand by all of the different labs.”
The recruitment drive has already paid dividends. Jason Wei, a key researcher behind OpenAI’s advanced models, recently joined Meta’s superintelligence team. Hyung Won Chung, another OpenAI researcher, is also making the switch.
Infrastructure investments dwarf previous spending
Meta’s commitment extends far beyond talent acquisition. The company is constructing massive data centers that will consume multiple gigawatts of power. These facilities represent the largest AI infrastructure investments in corporate history.
Zuckerberg revealed that Meta is building data centers covering “a significant portion of the footprint of Manhattan.” The Hyperion facility alone will scale to five gigawatts over the coming years.
“We can basically do this all funded from the cash flow of the company,” Zuckerberg said, highlighting Meta’s strong financial position.
A fundamental shift in AI strategy
Perhaps the most significant development is Meta’s potential abandonment of its open-source philosophy. For years, the company has freely shared its AI models and code with developers worldwide.
Now, internal discussions suggest Meta may keep its most powerful superintelligence models proprietary. This represents a dramatic strategic pivot driven by competitive pressures and performance concerns.
The company recently delayed releasing its “Behemoth” model after disappointing internal test results. This setback has prompted a serious reconsideration of Meta’s traditional approach to AI development.
Personal superintelligence: The ultimate goal
Meta’s vision differs fundamentally from competitors like OpenAI and Google. While other companies focus on automating economic productivity, Meta targets personal enhancement and lifestyle improvement.
“This year is going to set the course for the future,” Zuckerberg declared in February, calling artificial intelligence “potentially one of the most important innovations in history.”
The concept of personal superintelligence extends to Meta’s augmented reality ambitions. Zuckerberg predicts that future AI glasses will provide cognitive advantages to users. He envisions AI companions that can observe daily life and provide continuous assistance.
Industry disruption accelerates
Meta’s aggressive hiring strategy is forcing competitors to increase their compensation packages. The talent acquisition war has created unprecedented salary inflation across the tech industry.
Major players are responding with massive investments. Microsoft has committed over $13 billion to OpenAI. Amazon has invested $8 billion in Anthropic. Google paid $3 billion to license Character.AI’s technology and hire its team.
Internal challenges persist
Despite the ambitious plans, Meta’s AI division faces significant internal hurdles. Employee turnover has increased due to grueling development schedules and management conflicts.
Some team members are leaving for rival companies, frustrated by the intense work pace and internal politics. The departures highlight the challenges of managing such a high-pressure, high-stakes initiative.
The competitive landscape intensifies
The race for artificial general intelligence has become Silicon Valley’s defining battle. Companies are pursuing AGI—machines that can perform any intellectual task humans can do. Superintelligence would go even further, exceeding human cognitive capabilities.
Meta’s restructured approach under Wang’s leadership represents a complete reimagining of the company’s AI strategy. The new lab operates separately from other divisions, with direct access to Zuckerberg.
Regulatory considerations loom
Meta must navigate its investments carefully. The Federal Trade Commission recently challenged the company’s past acquisitions of Instagram and WhatsApp. The unusual structure of the Scale AI deal may help address some regulatory concerns.
The verdict remains uncertain
Whether Meta’s superintelligence gamble will succeed remains unclear. The company has assembled unprecedented resources, top talent, and infrastructure. However, the technical challenges of achieving superintelligence are immense.
The initiative represents Zuckerberg’s largest investment yet in transformative technology. Success could establish Meta as the dominant force in artificial intelligence. Failure could become another costly lesson in visionary ambition meeting technological reality.
Do you believe Meta can achieve superintelligence, or is this another expensive experiment? What impact could personal AI assistants have on daily life? Join the conversation and share your perspective on Meta’s latest ambitious venture.