Meta’s race toward superintelligence has entered a new phase. According to sources familiar with the matter, the social media giant is preparing to launch a dedicated “superintelligence lab” in early 2025, marking its most ambitious AI initiative to date.
The lab, which has reportedly been in stealth development for over a year, will focus exclusively on creating AI systems that surpass human capabilities across multiple domains simultaneously – the technical definition of artificial general intelligence (AGI) that has become Silicon Valley’s latest obsession.
This move positions Mark Zuckerberg’s company in direct competition with OpenAI, Anthropic, and Google DeepMind in the increasingly crowded race to build what some researchers call “god-like” AI systems. Industry insiders suggest the lab has already attracted over 40 top AI researchers, many poached from academic institutions with seven-figure compensation packages.
“Meta is essentially building a Manhattan Project for superintelligence,” said Rajiv Krishnan, venture partner at Threshold Ventures, who tracks AI investments. “The scale of computing resources they’re allocating to this is unprecedented outside of classified government projects.”
What separates this initiative from Meta’s existing AI work is its singular focus on systems that can autonomously improve themselves – a capability that both excites and terrifies experts. Unlike the company’s current large language models powering chatbots and content moderation, superintelligent systems could potentially solve scientific problems beyond human comprehension.
The financial commitment is equally staggering. Sources indicate Meta plans to invest upwards of $7 billion in specialized AI chips and computing infrastructure over the next three years, representing the largest single R&D bet in the company’s history.
A Meta spokesperson declined to confirm specific details but provided a statement: “We believe advanced AI will transform how people connect and experience the world. While we don’t comment on unannounced projects, we’re committed to developing AI responsibly and in ways that benefit humanity.”
The timing isn’t coincidental. Tech giants are locked in an increasingly expensive arms race to dominate what many believe will be the most transformative technology of the century. Just last quarter, Alphabet allocated $6.9 billion to its DeepMind division, while Microsoft has poured over $13 billion into its OpenAI partnership.
What makes Meta’s approach notable is its apparent willingness to pursue superintelligence without the research guardrails that competitors have at least publicly embraced. Internal documents reviewed by industry analysts suggest the company is prioritizing development speed over extensive safety protocols.
This strategy has raised eyebrows among AI safety advocates. “The problem with rushing toward superintelligence is that we don’t fully understand the implications,” explained Dr. Elena Morrow, director of the Technology Ethics Institute. “Meta appears to be taking a ‘build first, ask questions later’ approach that could have serious consequences.”
The company’s troubled history with data privacy and content moderation amplifies these concerns. Critics wonder whether Meta can be trusted with potentially the most powerful technology ever created when its existing platforms have struggled with basic harm prevention.
For Canadian businesses and investors, Meta’s superintelligence push creates both opportunities and challenges. Toronto’s Vector Institute has already seen three senior researchers depart for the project, continuing the brain drain that has plagued Canada’s AI sector for years.
“We’re watching our homegrown talent get scooped up by American tech giants with bottomless budgets,” said Jordan Chen, who leads a Toronto-based AI startup. “It’s impossible to compete with the salaries Meta is offering.”
At the same time, Meta’s aggressive push could accelerate the timeline for widely available AGI systems, potentially democratizing access to cognitive tools that would otherwise remain in the hands of a select few tech giants.
The financial markets have taken notice. Meta’s stock has climbed 12% since rumors of the lab first surfaced last month, adding nearly $150 billion to its market capitalization. Analysts at Morgan Stanley recently upgraded their target price, citing the company’s “underappreciated lead in foundational AI research.”
Behind closed doors, the project has reportedly created tensions within Meta itself. Several senior engineers have expressed concerns about rushing toward superintelligence without adequate safety measures, according to two current employees who requested anonymity.
“There’s a sense that we’re building something we don’t fully understand,” one source said. “Some teams are pushing for more robust containment protocols, but they’re being overruled by leadership eager to announce a breakthrough.”
The Canadian government has taken notice as well. Innovation Minister François-Philippe Champagne recently signaled that foreign investments in advanced AI could face increased scrutiny under national security provisions – a move widely interpreted as responding to concerns about superintelligence development.
For everyday users of Meta’s platforms, the immediate impact may be subtle. The company plans to gradually integrate capabilities from its superintelligence research into products like Instagram, WhatsApp, and its Reality Labs VR/AR initiatives. But the long-term implications could be profound, potentially transforming how we interact with technology in ways difficult to imagine today.
As 2025 approaches, the question isn’t just whether Meta can achieve superintelligence before its rivals, but whether any company should be pursuing such technology without broader societal consensus on governance and safety. The race is accelerating, and the finish line remains shrouded in both promise and peril.
