Part of: this research — view all 6 rows
Thinking Machines Lab (TML) launched with a heavy concentration of ex-OpenAI talent, but the verifiable count of qualifying senior staff is much smaller than its overall headcount. The strict resolution criteria require evidence of specific senior titles or leadership roles at OpenAI, Anthropic, or Google DeepMind. At present, a solid baseline of around 5 to 9 qualifiers can be confirmed. CEO Mira Murati (former OpenAI CTO) and Chief Scientist John Schulman (OpenAI co-founder and research lead) are near-certain qualifiers Thinking Machines Lab Former OpenAI CTO Mira Murati unveils Thinking .... Public profiles also confirm Alexander Kirillov (led OpenAI's multimodal group) https://alexander-kirillov.github.io/, Rowan Zellers (led GPT-4o video perception) Rowan Zellers' Homepage, and Stephen Roller (former Senior Staff Research Scientist at DeepMind) Stephen Roller. Because OpenAI largely uses a flat "Member of Technical Staff" title, other alumni only qualify if public sources verify they held specific lead roles.
A major constraint on TML's senior headcount is heavy and ongoing attrition. By mid-2026, roughly a third of the 42-person founding team had already departed https://www.businessinsider.com/thinking-machines-lab-key-talent-meta-openai-xai-2026-5 https://www.businessinsider.com/thinking-machines-lab-key-talent-meta-openai-xai-2026-5. Crucially, four of the six original cofounders have left, including Barret Zoph, Luke Metz, and Lilian Weng returning to OpenAI, and Andrew Tulloch moving to Meta. Meta and OpenAI have maintained steady poaching pressure Meta has hired five founding members of Mira Murati's .... As frontier labs accelerate their research—with OpenAI pushing toward recursive self-improvement and Anthropic reporting R&D progress acceleration—the gravitational pull of the leading labs may make it increasingly difficult for TML to retain its most senior AGI-focused researchers. Heavy attrition headwinds and intense financial retention efforts by frontier labs also make it harder for TML to recruit qualifying replacements, favoring a somewhat more cautious outlook than its resources alone would suggest . TML's current open-weights model, Inkling 2 Small, scores far behind frontier models on the Artificial Analysis index (30.6 vs. Claude Fable 5.2 at 64.6), which could further diminish its draw for top-tier talent.
Counterbalancing these departures are TML's massive resources, which provide substantial capacity for new senior hires. The lab is reportedly closing a $5B funding round at a $40B valuation Thinking Machines Lab Is Closing in on a $5 Billion New ... Thinking Machines Lab Is Closing in on a $5 Billion New ... and has secured a 1-gigawatt Vera Rubin compute partnership with Nvidia targeted for early 2027 deployment NVIDIA and Thinking Machines Lab Announce Long-Term .... The lab lists dozens of open positions https://jobs.ashbyhq.com/ThinkingMachines and explicitly targets senior engineers Technical Sourcer @ Thinking Machines Lab. These financial and computational resources should allow TML to backfill departures and recruit new senior researchers from Google DeepMind and OpenAI, but competition and retention incentives at those labs make roughly stable qualifying headcount more plausible than substantial net growth through August 2027. (Note that hires from non-qualifying labs, such as Meta's Soumith Chintala, do not count toward this total https://www.businessinsider.com/thinking-machines-lab-key-talent-meta-openai-xai-2026-5.)
The primary uncertainties lie in how successfully TML can leverage its capital against the prestige of accelerating frontier labs, and how leniently resolution criteria are applied to ambiguous public profiles. My median is 8 qualifying researchers, with the 25th and 75th percentiles at 6 and 12, respectively, and the 10th and 90th percentiles at 4.5 and 17.5. The lower bound accounts for continued correlated departures, such as another leadership exit or an acquisition, which Meta reportedly explored earlier Meta has hired five founding members of Mira Murati's ... Meta has hired five founding members of Mira Murati's .... The long right tail reflects a scenario where TML's new funding round and compute deployment trigger a major cluster of senior hires, or where a deep dive into public records reveals a larger-than-expected number of retained founding staff who demonstrably led projects at OpenAI.
Set against related questions, the distribution was lowered slightly to reflect heavy attrition headwinds and intense financial retention efforts by frontier labs .
Ask a followup
Sign in to run · $20 free credit, no card · every claim cited