Thinking Machines Lab Gains Key Meta Talent
Weiyao Wang worked at Meta for eight years straight out of college. He helped develop multimodal perception systems and worked on open-world segmentation efforts, such as SAM3D. Wang's last day at Meta came last week. Now he has moved to Thinking Machines Lab.
Thinking Machines Lab, an AI startup, grows across several areas. The company recently closed a multibillion-dollar cloud agreement with Google. That deal provides access to Nvidia's newest GB300 chips. TML counts among the initial startups using that hardware.
Google revealed the deal on Tuesday during its Cloud Next event. It builds on a prior collaboration with Nvidia. The setup places TML on infrastructure equal to Anthropic and Meta. Reports indicate Meta discussed buying Thinking Machines Lab about a year ago. Lately, Meta has recruited several of TML's founders.
Talent Shifts in Both Directions
Hiring trends stay active. Wang and Kenneth Li represent recent additions to TML. Li holds a Harvard PhD. He spent 10 months at Meta before starting at TML this month. Business Insider noted last week that Meta took seven founding members from TML. LinkedIn profiles reveal TML recruits more researchers from Meta than from any other company.
Soumith Chintala serves as TML's CTO. He spent 11 years at Meta and co-created PyTorch, an open-source framework for deep learning that supports much global AI work. Chintala departed Meta in late 2025. TML named him CTO earlier this year. Piotr Dollár, a Meta employee for 11 years, acted as research director. He co-wrote the Segment Anything model, now part of TML's technical team.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Andrea Madotto worked as a research scientist in Meta's FAIR group on multimodal language models. She joined TML in December. James Sun, a software engineer with almost nine years at Meta on LLM pre-training and post-training, also switched over.
Recruits from Across the Industry
TML pulls experts from other places too. Neal Wu won three gold medals at the International Olympiad in Informatics. He helped start the coding startup Cognition and came to TML early this year. Jeffrey Tao's path includes Waymo, Windsurf, and OpenAI.
Muhammad Maaz completed a research fellowship at Anthropic. Erik Wijmans came from Apple. Liliang Ren spent two and a half years on Microsoft's AI Superintelligence team. There he pre-trained OpenAI models for code. He arrived at TML in March.
The startup employs about 140 people now. Meta offers pay packages in the seven figures with few conditions. Researchers consider options carefully. Thinking Machines Lab holds a $12 billion valuation. That seems high for a firm with only one product released. Still, it trails the huge valuations of OpenAI and Anthropic, leaving room for growth.
A TML spokesperson contacted Friday morning offered no comment.
Thinking Machines Lab focuses on advanced AI systems. Meta, through its FAIR lab, leads in areas like computer vision and language models. PyTorch powers tools from research labs to production systems worldwide. The Segment Anything model advances image segmentation tasks. Such moves highlight competition for top AI minds amid rapid industry growth.