Meta once had some of the brightest minds in artificial intelligence. But over the past few years, many of those top researchers have walked out the door — building their own startups or joining rivals like OpenAI, Anthropic, Google, and Mistral. Now, Mark Zuckerberg is spending staggering amounts of money trying to bring that kind of talent back.
How Meta Lost Its Edge
Meta’s AI research lab, once considered world-class, has faced a steady “brain drain.” Key figures who worked on breakthroughs at Meta have since founded or joined leading AI companies such as Perplexity, Fireworks AI, and World Labs. Others were poached by OpenAI and Google.
Former employees told Forbes that Meta already had the best talent — but failed to keep them. As one ex-staffer put it: “This is Mark trying to undo the loss of talent.”
Zuckerberg’s Billion-Dollar Bids
In response, Zuckerberg has gone on a hiring spree, offering eye-popping salaries and benefits. Reports suggest he has dangled pay packages worth over $1 billion spread across multiple years to lure leading AI scientists from competitors. Meta has managed to poach some high-profile researchers, but many have turned him down, choosing companies where they see greater long-term impact and equity value.
Some critics compare Meta’s approach to a struggling sports team overpaying free agents. One AI startup founder told Forbes: “Meta is the Washington Commanders of tech companies. They massively overpay for okay-ish AI scientists, and people think they’re the best just because of the paycheck.”
Rivals Push Back
Other AI labs, like Anthropic, have openly rejected Meta’s huge offers. CEO Dario Amodei noted that pay inflation doesn’t necessarily make sense: “If Zuckerberg throws a dart and hits your name, that doesn’t mean you should make 10 times more than the person next to you with the same skills.”
Retention numbers back that up. According to VC firm SignalFire, Anthropic keeps about 80% of its staff, compared with DeepMind at 78%, OpenAI at 67% — and Meta trailing at 64%.
A Culture of Chaos
Beyond the money, Meta has also been criticized for its chaotic culture. Researchers describe constant leadership changes, shifting priorities, and clashing teams. AI projects were rushed, deadlines were tight, and some teams were reassigned to work on the Metaverse even as the AI boom took off.
One former researcher summed it up: “Llama 4 was a disaster.” Another described Meta’s AI environment as “disheartening, overworked, and confusing.”
What’s Next for Meta’s AI Gamble?
Despite the setbacks, Zuckerberg isn’t slowing down. He has hired big names like Alexandr Wang (former Scale AI CEO) and Daniel Gross (cofounder of Safe Superintelligence) to lead new AI efforts at Meta. The company has also launched a new “superintelligence” unit to push ahead in building next-generation AI systems.
Still, competition is fierce. Rivals like OpenAI, Anthropic, Google DeepMind, Mistral, and Elon Musk’s xAI are attracting top talent without needing to pay “insane” salaries. For Zuckerberg, the challenge isn’t just writing bigger checks — it’s convincing researchers that Meta is the place to do their life’s best work.
#ArtificialIntelligence #Meta #Zuckerberg #TechNews #AIResearch