Zuckerberg’s Poaching Top OpenAI Talent for Meta Super Intelligence Labs

ark Zuckerberg’s Bold Move: Meta’s Super Intelligence Team and the AI Talent WarIn a stunning display of ambition, Mark Zuckerberg, CEO of Meta Platforms, has gone all-in on building Meta’s AI dominance, assembling a dream team of researchers poached from rivals like OpenAI, Anthropic, and Google.

Meta’s newly unveiled Meta Super Intelligence Labs (MSL) is a powerhouse of talent, signaling Zuckerberg’s relentless pursuit of superintelligence. “As the pace of AI progress accelerates, developing superintelligence is coming into sight,”

Zuckerberg declared in an internal memo. “I believe this will be the beginning of a new era for humanity, and I am fully committed to doing what it takes for Meta to lead the way.”

Here’s how this high-stakes AI talent war unfolded and what it means for the future of artificial intelligence.

The Catalyst: Llama 4 and Meta’s AI AmbitionsA few months ago, Meta released Llama 4 in small and medium sizes, with the large model still under wraps.

While the models were impressive, they weren’t enough to position Meta as the leader in the global AI race, trailing competitors like OpenAI and Google. To win, Meta needed the best models on the planet—and Zuckerberg knew it.

Recognizing that Meta’s existing efforts weren’t cutting it, he took decisive action.

Zuckerberg crafted a secret list of the world’s top AI researchers and launched an aggressive campaign to recruit them, reportedly offering $100 million signing bonuses and even higher pay packages, as confirmed by OpenAI CEO Sam Altman on a podcast with his brother.

This wasn’t just a hiring spree—it was a calculated move to reshape the AI landscape.The $14 Billion Scale AI Deal: A Strategic MasterstrokeMeta’s acquisition of a 49% stake in Scale AI for $14.3 billion was a pivotal moment.

Scale AI, known for its data collection and labeling infrastructure for large language models, brought more than just technology to the table.

The real prize was its team, led by Alexandr Wang, a prodigy hailed as one of the best CEOs of his generation.

Why didn’t Meta buy Scale AI outright? Two reasons: speed and regulatory hurdles. A full acquisition would have triggered lengthy reviews by U.S. and international regulators.

With the Trump administration now easing antitrust concerns, Meta opted for a minority stake to close the deal quickly. However, this move came at a cost: competitors like OpenAI and Google canceled their contracts with Scale AI, wary of sharing data with a Meta-backed entity.

But Zuckerberg had already secured what he needed: Scale’s data, infrastructure, and, most crucially, its leadership.

Building the Super Intelligence TeamAlexandr Wang, now Meta’s Chief AI Officer and head of MSL, was tasked with leading the charge toward superintelligence, alongside Nat Friedman, former GitHub CEO, who focuses on AI products and applied research.

Together, they spearheaded an aggressive poaching campaign that shook the AI industry.

Meta’s recruitment efforts hit OpenAI particularly hard, with the company losing at least eight top researchers, including the entire Zurich office, which housed talent instrumental in developing ChatGPT, reasoning models, and more.

Here’s a look at some of the all-star researchers now at Meta, as listed in Zuckerberg’s internal memo:

Trapit Bansal: Pioneered reinforcement learning and chain-of-thought reasoning, co-creator of OpenAI’s O-series models.

Shuchao Bi: Co-creator of GPT-4o Voice Mode and GPT-4o Mini, previously led multimodal post-training at OpenAI.

Huiwen Chang: Co-creator of GPT-4o’s image generation capabilities.

Ji Lin: Contributed to GPT-4o, O3/O4 Mini, and operator reasoning stacks.

Hongyu Ren: Co-creator of GPT-4o, O3/O4 Mini, and O1 Mini.

Johan Schalkwyk: Former Google Fellow, early contributor to Sesame, and technical lead for Maya.

Pei Sun: Led post-training, coding, and reasoning for Google DeepMind’s Gemini.

Jiahui Yu: Co-creator of GPT-4o, O3/O4 Mini, and O1.

Shengjia Zhao: Co-creator of ChatGPT, GPT-4, and led synthetic data efforts at OpenAI.

Joel Pobar: Inference expert from Anthropic, with prior experience at Meta.

This roster, part of a 50-member elite unit, represents a who’s who of AI innovation, with expertise spanning reinforcement learning, multimodal AI, and reasoning models.

Their mission? To build the next generation of AI capable of achieving superintelligence, aligning with Zuckerberg’s vision of a “new era for humanity.”

OpenAI’s Response: A Company in Crisis

OpenAI was caught off-guard by Meta’s aggressive tactics. As researchers jumped ship, OpenAI’s Chief Scientist, Mark Chen, issued an internal memo (leaked on Saturday) acknowledging the “massive brain drain.”

Describing the poaching as akin to “someone breaking into our home,” Chen vowed to fight back by recalibrating compensation and offering creative incentives to retain talent.

However, OpenAI faces a steep challenge: Meta’s trillion-dollar market cap and vast cash reserves dwarf OpenAI’s resources, even as the world’s largest private company.

Chen also admitted that OpenAI had been too focused on incremental, buzzworthy releases rather than the ultimate goal of superintelligence.

Sam Altman, reflecting on the broader AI race, emphasized OpenAI’s mission-driven approach, stating, “And maybe more importantly than that, we actually care about building AGI in a good way. Other companies care more about this as an instrumental goal to some other mission.”

With Meta capitalizing on OpenAI’s week-long break to pressure researchers with “exploding offers,” Chen urged his team to resist hasty decisions and reach out for support.

Meta’s Vision: The Meta Super Intelligence Labs (MSL)Zuckerberg formally introduced MSL in an internal memo obtained by WIRED and Bloomberg.

“We’re going to call our overall organization Meta Super Intelligence Labs (MSL),” Zuckerberg wrote.

“This includes all of our foundations, product, and FAIR teams, as well as a new lab focused on developing the next generation of our models.” The organization consolidates Meta’s AI efforts, including its foundational models (like Llama), product teams, and the former FAIR (Facebook AI Research) team, under a unified mission: to develop superintelligent AI.

The memo highlighted the new lab’s focus on next-generation models, with Wang and Friedman at the helm. Notably, Yann LeCun, Meta’s long-time AI chief, was not mentioned in the MSL leadership, suggesting a shift in priorities toward Wang’s vision.

Zuckerberg’s memo emphasized Meta’s unique position to deliver superintelligence, citing its massive investments in data centers and computing power. MSL aims to integrate advanced AI into Meta’s products like Facebook, Instagram, and WhatsApp, while also pushing the boundaries of artificial general intelligence (AGI).

The Bigger Picture: A New Era in AI

Meta’s bold moves—poaching top talent, acquiring Scale Ascendancy, and launching MSL—signal a seismic shift in the AI landscape.

Zuckerberg’s singular focus on superintelligence has disrupted the industry, forcing competitors like OpenAI to rethink their strategies. While OpenAI scrambles to retain talent and refocus on long-term goals, Meta is leveraging its financial might and strategic acquisitions to accelerate its AI ambitions.The stakes couldn’t be higher.

Superintelligence—the ability for AI to surpass human intelligence across all domains—remains the ultimate prize. Meta’s all-star team, backed by unparalleled resources, is now working around the clock to make that vision a reality.

However, the aggressive poaching tactics have raised ethical questions about talent acquisition in the AI race, with some critics arguing that Meta’s moves could destabilize smaller competitors.

https://www.wired.com/story/mark-zuckerberg-welcomes-superintelligence-team/