Apple’s lead robotics AI researcher joins Meta

Apple has lost several key AI researchers to rival companies, with lead robotics researcher Jian Zhang joining Meta’s Robotics Studio, the company confirmed.

Three other AI researchers from Apple’s large language models team—John Peebles, Nan Du, and Zhao Meng—are also departing, with Peebles and Du headed to OpenAI, and Zhao joining Anthropic.

These exits come after a series of recent departures from Apple’s Foundation Models team, which played a central role in building the Apple Intelligence platform.

Meta and other AI firms have attracted Apple employees with higher compensation, while Apple is reportedly considering greater reliance on external technology for its AI projects.

Apple’s share price fell 1.5% after news of the departures.

Zhang led a team focused on automation within Apple’s AI and machine learning division, which is distinct from Apple’s robotics product development group.

Several more Apple AI staff are reportedly interviewing elsewhere amid concerns about internal morale and strategy shifts.

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ AI talent wars reveal unsustainable compensation inflation

Apple’s recent exodus highlights how extreme financial competition for AI talent is creating market distortions that may not be sustainable long-term.

Meta’s recruitment strategy has included offering packages like the $200 million multiyear deal to Ruoming Pang, Apple’s former models team leader1.

This represents compensation levels far beyond traditional tech industry standards, even for senior executives.

However, this aggressive poaching strategy appears to have retention challenges, as several of Meta’s recently hired AI researchers have already quit, according to industry reports1.

The pattern suggests that while massive compensation can attract talent initially, it doesn’t guarantee long-term retention or productivity, potentially creating expensive talent churn cycles.

With global demand for AI professionals vastly outstripping supply—an estimated 400,000 ML developers needed versus only 92,000 currently in Global 500 companies2—companies may be overpaying for talent without proportional returns on investment.