Lilian Weng pivots from Thinking Machines startup to OpenAI return

Weng previously served as the VP of AI Safety Research at OpenAI before her stint at the startup. In her new capacity, she will focus on accelerating internal research, specifically targeting recursive self-improvement—a process enabling AI systems to iteratively enhance their own capabilities. This shift back to a large-scale organization may offer a reprieve from the intense, constant pressure she identified as the primary driver for her initial resignation.

Mira Murati, a co-founder of Thinking Machines and former OpenAI CTO, publicly supported Weng’s decision to prioritize her health, though it remains unclear if she was privy to Weng’s imminent return to their former employer. In a competitive landscape where AI labs aggressively hunt for top-tier talent, the transition underscores the difficulty of balancing high-stakes startup culture with personal sustainability.