Meta's AI Superintelligence Lab Hit by Immediate Exodus

In a stunning blow to Meta's artificial intelligence ambitions, at least three high-profile researchers recruited to its newly formed Superintelligence Lab have resigned just weeks after joining. According to a WIRED exclusive report by Zoë Schiffer and Will Knight, two of these departing scientists have already returned to OpenAI – their previous employer – after less than a month-long stint at Meta. This rapid reversal comes shortly after CEO Mark Zuckerberg announced the lab with fanfare, promising to build artificial general intelligence (AGI) and offering compensation packages rumored to reach into the tens of millions of dollars for top talent.

Meta's Superintelligence Lab, launched with ambitious goals, is already seeing key recruits depart. (Credit: WIRED)

The Sting of Short Tenures and Swift Returns

The speed of the departures is particularly damaging. Researchers typically endure lengthy non-compete periods when leaving AI giants like OpenAI. The fact that these individuals returned almost immediately suggests their Meta experience was profoundly negative or that OpenAI aggressively facilitated their return. Sources indicate the issue wasn't solely financial – Meta is known to have offered staggering sums, potentially exceeding $100 million in total compensation for some recruits – but rather involved deeper concerns about mission alignment and leadership dynamics.

"Mark Zuckerberg just went on this massive multi-billion dollar recruiting spree. He's giving people pay packages more often associated with professional sports stars," noted Zoë Schiffer on WIRED's Uncanny Valley podcast. "The fact that some people are already leaving very, very shortly after starting... is just such a bad sign. It just seems like there's a lot of chaos right from the beginning."

Leadership, Mission, and the Allure of OpenAI

A critical factor appears to be the perceived contrast in leadership and purpose between Meta and rivals like OpenAI. While Meta AI chief Yann LeCun is a respected pioneer, the operational leadership of the Superintelligence Lab falls to Alexander Wang, CEO of Scale AI. Wang, though successful, is perceived by some in the research community as less technically immersed than leaders at competing AI labs. This contrasts sharply with OpenAI's Sam Altman and especially former CTO Mira Murati, who, while not hardcore researchers themselves, cultivated significant loyalty and a sense of shared, world-changing mission.

"You'll hear people say... [Alexander Wang] is not a hardcore researcher... Sam Altman's not a super technical researcher either, but he really has almost what I would say is a cult of personality around him at OpenAI," Schiffer explained. "Mira Murati... was able to talk to researchers. They really felt understood by her. They trusted her."

The fundamental mission also diverges. Where OpenAI and Anthropic frame their work in terms of transformative potential (biomedical breakthroughs, personalized medicine), Meta's focus is often perceived internally as enhancing its core social media products – "AI slop for Instagram and Facebook," as one source characterized it to Schiffer. While potentially lucrative, this application-focused goal may lack the inspirational pull for researchers drawn to the loftier promises of AGI.

Echoes in the Job Market: AI Eliminates Entry-Level Roles

This talent turmoil coincides with broader shifts in the tech labor market driven by AI. A separate WIRED report by Will Knight detailed a Stanford University study providing strong evidence that AI is actively eliminating entry-level positions, particularly in software development and customer service. The study found a 16% decline in employment for workers aged 22-25 in these sectors.

AI's impact extends beyond research labs, reshaping entry-level job markets. (Credit: WIRED)

This trend raises profound concerns beyond immediate job loss. Entry-level roles traditionally serve as crucial training grounds. "The main thing that you do in these entry-level jobs is figure out how to be a real human and how to live in an adult world," commented WIRED Senior Politics Editor Leah Feiger on the podcast. "We're losing a big pipeline... How do you get experience?" Companies increasingly seek mid-level engineers adept at collaborating with AI, but the pathway to gaining that essential experience is vanishing for new graduates.

Musk's Legal Gambit and the Shifting Landscape

Further illustrating the industry's volatility, Elon Musk's xAI filed a lawsuit against Apple and OpenAI, accusing them of anti-competitive behavior related to App Store rankings and integration. While this feud has deep personal roots stemming from Musk's co-founding and subsequent departure from OpenAI, it underscores the fiercely competitive and litigious environment surrounding AI platform dominance.

An Unsustainable War for Minds?

Meta's early stumbles with its Superintelligence Lab, coupled with the erosion of entry-level tech jobs, paint a picture of an AI industry at an inflection point. The astronomical sums being thrown at top researchers may not be sustainable or sufficient if fundamental issues of culture, leadership, and mission aren't addressed. Simultaneously, the automation of junior roles threatens the long-term talent pipeline essential for innovation. The race for AI supremacy isn't just about building smarter models; it's about building organizations capable of attracting, retaining, and nurturing the human intelligence required to guide this powerful technology. Meta's brain drain is a stark warning that money alone can't buy loyalty or solve deeper strategic challenges in this high-stakes arena.

Source: Reporting based on WIRED's Uncanny Valley podcast episode and exclusive articles by Zoë Schiffer and Will Knight: Researchers Are Already Leaving Meta’s New Superintelligence Lab, AI Is Eliminating Jobs for Younger Workers.