This is why Meta fired employees in its AI division
What's the story
Meta's former research scientist, Tian Yuandong, has revealed that the company's restructuring of its artificial intelligence (AI) team was driven by limited computing resources. The revelation comes as a reason behind the recent layoffs at Meta's AI research unit. Yuandong, who previously served as a research scientist director at Meta FAIR (Facebook's Fundamental AI Research team), made these comments during an interview on Silicon Valley 101.
Resource constraints
Internal conflicts at Meta's AI research unit
Yuandong said that the growing focus on large language models (LLMs) in global AI development has led to internal conflicts within Meta's AI research unit. The limited computing resources have been a major factor in these tensions. This explanation sheds light on why Meta recently laid off around 600 employees from its AI research unit, including Yuandong himself.
Strategy shift
Layoffs and strategy shift at Meta
The layoffs came months after Meta changed its strategy through the acquisition of a data-annotation firm and hiring talent from rival AI labs. Despite the upheaval, the newly formed Super Intelligence Lab remained largely untouched. It is led by the company's new Chief AI Officer Alexandr Wang, the ex-CEO and founder of Scale AI.
Profile
A look at Yuandong's work
Working at Meta for over 10 years, Tian holds degrees from China's Shanghai Jiao Tong University and Carnegie Mellon University in the US. He has also done extensive research on LLMs and reinforcement learning. While a number of Big Tech companies and start-ups have reached out to him, Tian is yet to decide on his next course of action. He hopes his next work will combine frontier research with application.