Meta is placing a massive wager on the future of artificial intelligence with a planned multibillion dollar investment in Scale AI.
This fresh alliance will give Meta a near half share in Scale, and CEO Alexandr Wang is expected to step directly into the fray to lead a brand new ambitious lab inside the company — one focused on superintelligence.
Back when Meta paid extraordinary sums for WhatsApp and Instagram, skepticism reigned. Many wondered whether Mark Zuckerberg’s empire was just burning through money, but both platforms are now core elements of Meta’s dominance.
This time, the skepticism surrounds not a chat app or visual platform but data itself — specifically, the lifeblood that powers advanced AI systems.
Over recent years, industry leaders like OpenAI have built model after model on meticulously labeled datasets provided by firms like Scale. Now, these data providers are upping the ante, snapping up PhDs and senior technologists to push quality even higher.
Inside the High-Stakes Gamble
A source with knowledge of Meta’s recent struggles said the tech giant has grown frustrated by a lack of creative progress within its AI teams. When Meta’s much-hyped Llama 4 failed to outshine models from rivals in China, the sense of urgency only grew sharper.
Making matters harder, Meta has recently seen notable AI talent walk out the door, with more than four percent of high performers jumping ship in 2024 alone.
So it is not just Scale’s data Meta is after, but also Wang himself. The thirtysomething founder has become something of a rising star among Silicon Valley’s elite — he is known for his magnetism, networking skills, and drive.
Even so, Wang’s resume is different from that of peers steering other major AI labs, who often lean heavily on research credentials. That may explain why, alongside Wang, Meta is said to be eyeing figures like Jack Rae from DeepMind to strengthen its research muscle.
The future for Scale as a company remains murky. While some AI labs are now building their own data operations internally, others chase new approaches centered on synthetic data generated by machines rather than people.
Earlier this year, reports surfaced that Scale AI failed to hit some revenue targets. Meanwhile, cutting-edge AI companies have been engaged in a relentless search for more powerful, cost-effective methods to find and refine data.
Industry insiders suggest that clients, wary of Scale’s new close ties to Meta, may soon look elsewhere for a more impartial partner. Already, competitors like Turing and Surge AI have noticed increased interest. As Turing’s Jonathan Siddharth put it, some clients simply prefer working with a company that is not in anyone’s corner.
How all this will shake out for Meta — and whether this bold acquisition will once again reshape the landscape — remains to be seen. Rival firms are not standing still, either. OpenAI is preparing its most advanced model yet, hoping to leapfrog Meta’s current performance and push the field’s boundaries even further.
