Meta is Betting Big on AI “Superintelligence” and Alexandr Wang

Meta is doubling down on artificial intelligence again. In its latest high-stakes move to keep pace with OpenAI, Google, and Anthropic, Meta is launching a new “superintelligence” AI research lab led by Scale AI founder Alexandr Wang. The goal? Develop AI so advanced it could outthink humanity itself, or at least help Meta claw its way back to the front of the AI race. We just reported on Meta’s multibillion-dollar investment into Scale AI. This bold partnership signals a serious shift in Meta’s AI strategy, and possibly its internal power structure.

What is superintelligence?

Now, what exactly is Meta superintelligence? Good question. Nobody really knows. It’s the vague holy grail of AI: a system smarter than all of humanity combined. A few years ago, this would’ve sounded like a sci-fi fever dream. Today? It’s a recruiting strategy.

Let’s be clear: this isn’t just about vision. This is Meta trying to prove it can still compete with OpenAI, Google DeepMind, and Anthropic.

Meta’s had a rough couple of years on the AI front: multiple reorganizations, brain drain from its research teams, and high-profile departures. Despite offering eye-watering compensation packages (some reportedly in the nine-figure range), they’re still losing top AI talent to the competition.

And the internal dynamics? Complicated. Meta already has FAIR (Fundamental AI Research), the lab led by longtime Meta chief scientist Yann LeCun. But LeCun has always been a bit of an AI purist, not exactly onboard with the AGI/superintelligence hype train.

He believes LLMs like GPT or Llama aren’t the path to human-level intelligence, and he’s been pretty vocal about it. Which raises the question: what happens when Meta throws its weight behind a totally different vision under Wang? Are we watching a shift in power, or just more internal chaos?

The quasi-acquisition strategy

This move also highlights a growing trend in Big Tech: the “quasi-acquisition.” Microsoft did it with Inflection. Amazon did it with Adept. Google did it with Character AI.

These deals look a lot like talent raids with investment frosting. They stop short of a full acquisition, but they still scoop up the brains, IP, and momentum — and conveniently avoid pesky antitrust issues.

The stakes are high for Meta. While the open-source Llama models gave them a brief moment of AI glory in 2023 and 2024, the recent release of Llama 4 was… well, a flop.

Critics called it rushed, overhyped, and lacking transparency. Meanwhile, Chinese open-source contenders like DeepSeek are starting to gain serious traction. The clock is ticking.

Wang, on the other hand, is having a moment. Scale AI built its empire by doing the grunt work that fuels modern AI: labeling data at scale. It may not be glamorous, but it’s essential. Wang has always argued that data is not commoditized, that future leaps in AI performance will come from smarter, better, and faster data pipelines. And now Meta is banking on that belief in a big way.

Will this deal turn Meta’s AI fortunes around? Too soon to tell. But one thing is clear: Zuckerberg is betting that Alexandr Wang is his ace. And if the superintelligence moonshot pays off, it won’t just be a win for Meta — it could rewrite the next chapter of the AI race.

But if not? At least Wang got the better end of the deal.

---

Remember, AI won’t take your job. Someone who knows how to use AI will. Upskilling your team today, ensures success tomorrow. In-person and virtual training workshops are available. Or, schedule a session for a comprehensive AI Transformation strategic roadmap to ensure your marketing team utilizes the right AI tech stack for your needs. From custom prompt libraries to AISO, Human Driven AI is your partner in AI success.