Andrew Tulloch, an AI researcher who was paid a reported $1.5 billion package over six years at Meta, is leaving the company. He'd been waiting to depart until Meta launched its new open-source AI models and Muse assistant, which dropped Tuesday. Tulloch came to Meta from Thinking Machine Labs and worked in the company's TBD lab, the core of Mark Zuckerberg's AI strategy run by Scale AI founder Alexandr Wang. The exact reasons for his exit aren't clear, and whether he'll continue AI research elsewhere remains unknown. The $1.5 billion figure, though disputed, would make Tulloch the highest-paid tech employee ever. That kind of compensation reflects how desperately companies compete for elite AI talent. Researchers at this level move between OpenAI, Anthropic, Google, Meta, and other labs based on opportunity and money. The field is hot enough that top people can either command massive salaries or raise billions for their own startups—Ilya Sutskever did it at OpenAI, Jeff Dean at Google. The real question is whether this talent arms race can sustain itself. Right now, a small number of world-class researchers have enormous leverage because supply can't keep up with demand. Eventually the market has to adjust, either through training more talent or through companies accepting that they can't all hire the absolute best. How long that takes is anyone's guess.