The meta’s loss is the thinking machine’s gain.

Weiyao Wang spent eight years at Meta, his first job out of college, where he helped build multimodal perception systems and contributed to open-world segmentation projects, including SAM3D. His last day at Meta was last week, and he has since joined the Thinking Machines Lab (TML).

His move to TML comes as the AI ​​startup expands across multiple areas. The company just signed a multibillion-dollar cloud deal with Google, giving it access to Nvidia’s latest GB300 chip, making it one of the first startups to run on this hardware.

The deal, announced at Google Cloud Next this Tuesday, follows a previous partnership with Nvidia that puts TML on the same infrastructure tier as Anthropic and Meta. (Meta was reportedly in talks to acquire Thinking Machines this time last year, and recently eliminated TML’s founders one by one.)

The image of human resources remains in flux. Wang and Kenneth Li, a Harvard doctoral student who spent 10 months at Meta before joining TML this month, are the latest example of talent acquisition going both ways. Business Insider reported last week that Meta has now poached seven founding members of TML. A look back at recent hires shows that Thinking Machines are once again raiding the meta. At least based on reviews of LinkedIn profiles, TML appears to hire more researchers from Meta than any other single employer.

The most prominent is Soumith Chintala, CTO of TML. He spent 11 years at Meta and co-founded PyTorch, an open source deep learning framework that currently powers most of the world’s AI research. He will retire from Meta at the end of 2025 and was appointed CTO earlier this year. Piotr Dollár, another 11-year meta veteran who served as research director and co-authored the influential Segment Anything model, is currently on TML’s technical staff. Andrea Madotto, a researcher in Meta’s FAIR department specializing in multimodal language models, joined TML in December. James Sun, a software engineer who has been working on LLM pre- and post-training at Meta for nearly nine years, also made the jump.

TML draws talent from outside the meta as well. Neal Wu, three-time International Information Olympiad gold medalist and founding member of buzzy coding startup Cognition, joined earlier this year. Jeffrey Tao came to us via Waymo, Windsurf, and OpenAI. Muhammad Mars previously held a research fellowship at Anthropic. Erik Wijmans from Apple has arrived. Before joining the company in March, Liliang Ren spent two and a half years pre-training code for OpenAI models on Microsoft’s AI Superintelligence team.

The startup currently has around 140 employees.

Meta’s salary package (seven figures, no strings attached) is well known by now. For researchers weighing other options, the calculation could be as simple as this. Thinking Machines Lab is currently worth $12 billion. While these numbers are unimaginable for a company at this stage of its technology cycle (the company has only released one product to date), there is still significant room for financial improvement compared to OpenAI and Anthropic’s record valuations.

A TML spokesperson, reached on Friday morning, declined to comment on the matter.