Meta’s loss is Thinking Machines’ gain: Startups — Key Details

AI startup Thinking Machines Lab (TML) is rapidly expanding its formidable talent pool, notably attracting key researchers from tech giant Meta, even as the two companies engage in a reciprocal talent exchange. The most recent high-profile departure from Meta is Weiyao Wang, an eight-year veteran specializing in multimodal perception systems, who joined TML last week. This talent influx coincides with TML's recent multibillion-dollar cloud deal with Google, securing access to Nvidia’s cutting-edge GB300 chips, further cementing its position as a major player in the artificial intelligence landscape.

The movement of talent between Meta and TML has become a prominent feature of the competitive AI ecosystem. While Business Insider recently reported that Meta has successfully recruited seven of TML's founding members, TML has been actively — and seemingly more successfully — recruiting from Meta. A review of recent hires suggests that TML has drawn more researchers from Meta than from any other single employer, indicating a significant strategic advantage in talent acquisition.

Among the most impactful additions is Soumith Chintala, TML’s Chief Technology Officer, who departed Meta in late 2025 after an 11-year tenure. Chintala is widely recognized as a co-founder of PyTorch, the open-source deep learning framework that forms the foundation of much of today's AI research. Another significant hire is Piotr Dollár, an 11-year Meta veteran who previously served as a research director and co-authored the influential Segment Anything model, now part of TML’s technical staff.

Further bolstering TML’s expertise are Andrea Madotto, a research scientist from Meta’s FAIR division with a focus on multimodal language models, who joined in December, and James Sun, a software engineer with nearly nine years at Meta specializing in large language model pre- and post-training. These additions highlight TML's concentrated effort to integrate top-tier talent with deep experience in foundational AI technologies.

TML's recruitment strategy extends beyond Meta, attracting talent from across the industry's leading innovators. Neal Wu, a three-time gold medalist at the International Olympiad in Informatics and a founding member of the highly-regarded coding startup Cognition, joined TML early this year. Jeffrey Tao brings experience from Waymo, Windsurf, and OpenAI, while Muhammad Maaz previously held a research fellowship at Anthropic. Erik Wijmans arrived from Apple, and Liliang Ren joined in March after spending two and a half years on Microsoft’s AI Superintelligence team, where he focused on pre-training OpenAI models for code. The startup now boasts a headcount of approximately 140 employees.

TML's aggressive talent acquisition is complemented by robust strategic partnerships. The recently announced multibillion-dollar deal with Google Cloud, unveiled at Google Cloud Next, ensures TML gains early access to Nvidia's next-generation GB300 chips. This places TML in an elite group of AI companies, including Anthropic and Meta, that have secured premium infrastructure. This agreement follows an earlier partnership with Nvidia, underscoring TML's commitment to leveraging cutting-edge hardware for its AI development.

The startup’s rapid ascent is reflected in its staggering $12 billion valuation in a seed round. While such a figure would have been unprecedented in prior tech cycles for a company at this stage—especially one that has only released a single product, Tinker—it highlights the extraordinary valuations currently observed within the AI sector. This valuation, though substantial, still presents significant financial upside for new hires when compared to the even higher valuations of industry leaders like OpenAI and Anthropic.

For top AI researchers, the decision to join TML involves a careful consideration of financial incentives and growth potential. While Meta is well-known for offering lucrative seven-figure pay packages with no strings attached, TML presents a compelling alternative: the opportunity for substantial equity growth in a rapidly appreciating startup. The promise of being part of a company valued at $12 billion, with potential for further exponential growth in the burgeoning AI market, offers a different kind of reward to those weighing their career options. This dynamic competition for talent underscores the high-stakes environment in the race for AI dominance.

FAQ

Q: What is the core dynamic of talent movement between Meta and Thinking Machines Lab?

A: The movement is characterized as a "two-way street." While Meta has reportedly poached seven of Thinking Machines Lab's (TML) founding members, TML has also actively recruited numerous researchers from Meta, seemingly more than from any other single employer. This indicates intense competition for top AI talent between the two entities.

Q: Who are some of the key individuals joining Thinking Machines Lab, and what is their background?

A: Recent prominent hires at TML include Weiyao Wang, an eight-year Meta veteran specializing in multimodal perception systems, and Kenneth Li, a Harvard PhD who spent ten months at Meta. TML's CTO, Soumith Chintala, spent 11 years at Meta and co-founded PyTorch. Other Meta alumni include Piotr Dollár (research director, co-authored Segment Anything model), Andrea Madotto (research scientist, multimodal language models), and James Sun (software engineer, LLM pre- and post-training). TML has also attracted talent from other major firms like Cognition, Waymo, OpenAI, Anthropic, Apple, and Microsoft.

Q: What strategic advantages has Thinking Machines Lab recently secured?

A: Thinking Machines Lab recently signed a multibillion-dollar cloud deal with Google, granting it access to Nvidia's latest GB300 chips. This places TML in the same elite infrastructure tier as companies like Anthropic and Meta. Additionally, the startup boasts a $12 billion valuation in its seed round, which, despite having only released one product (Tinker), offers significant financial upside for researchers considering joining.