In a significant development shaking the artificial intelligence sector, Mira Murati’s high-profile startup, Thinking Machines Lab, is confronting a major talent drain as two of its founding members depart to rejoin industry titan OpenAI. This strategic move, confirmed through social media announcements on Wednesday, underscores the intense and ongoing competition for elite AI researchers and the formidable challenges even well-funded startups face in retaining top leadership.
Thinking Machines Lab Confirms Leadership Shakeup
Mira Murati, the former OpenAI CTO who launched Thinking Machines Lab in late 2024, publicly announced the departure of Barret Zoph, the company’s co-founder and Chief Technology Officer. In a concise post on the social media platform X, Murati stated, “We have parted ways with Barret.” She simultaneously introduced Soumith Chintala as the new CTO, praising him as a “brilliant and seasoned leader” with over a decade of contributions to the AI field. Notably, Murati’s announcement did not mention the simultaneous departure of co-founder Luke Metz or another former OpenAI staffer, Sam Schoenholz.
However, just 58 minutes later, Fidji Simo, OpenAI’s CEO of Applications, revealed the broader scope of the talent movement. Simo’s post welcomed “Barret Zoph, Luke Metz, and Sam Schoenholz back to OpenAI,” indicating the moves had been planned for several weeks. This rapid sequence of announcements highlighted the interconnected nature of the AI talent pool and the swift circulation of personnel among leading entities.
The Returning Talent: A Deep Bench of Experience
The individuals returning to OpenAI bring substantial experience. Barret Zoph previously served as OpenAI’s Vice President of Research and spent six years as a research scientist at Google. Luke Metz, a co-founder of Thinking Machines, had a multi-year tenure on OpenAI’s technical staff. Sam Schoenholz, whose LinkedIn profile still listed him at Thinking Machines at the time of the announcement, also has a prior history with OpenAI. This collective return represents a significant reacquisition of institutional knowledge and technical expertise for OpenAI.
Contextualizing the AI Talent War
While talent movement between tech giants and startups is a hallmark of Silicon Valley, the departure of multiple co-founders from a startup less than a year after its founding is particularly notable. Thinking Machines Lab, despite its impressive launch and funding, now joins a list of AI ventures navigating the gravitational pull exerted by established players with vast resources.
- •High-Profile Founding: Murati launched Thinking Machines with Zoph and Metz after leaving her role as OpenAI CTO in September 2024.
- •Substantial Backing: The startup secured a massive $2 billion seed round in July, led by Andreessen Horowitz with participation from Accel, Nvidia, AMD, and Jane Street, achieving a $12 billion valuation.
- •Previous Departures: This is not the first exit; co-founder Andrew Tulloch left to join Meta in October.
The loss of two co-founders, especially the CTO, could be perceived as a meaningful operational and symbolic setback. Thinking Machines had assembled a team of researchers from OpenAI, Meta, and Mistral AI, aiming to compete at the frontier of AI development.
OpenAI’s Revolving Door and Competitive Landscape
OpenAI itself has experienced notable departures of co-founders and key staff to competing ventures, illustrating the fluid dynamics of the industry. For instance, co-founder John Schulman left for Anthropic in August 2024 before briefly joining Thinking Machines Lab as Chief Scientist at its launch in February 2025. This pattern shows that talent circulation is multidirectional, though the net flow often favors organizations with proven scale and stability.
The competition extends beyond personnel. It encompasses compute resources, research breakthroughs, and product deployment. Startups like Thinking Machines, even with billions in funding, must execute flawlessly to compete with the infrastructure and data advantages of incumbents. Leadership stability is a critical component of that execution.
Strategic Implications for Thinking Machines Lab
For Thinking Machines, the immediate challenge is a leadership transition. The appointment of Soumith Chintala, a respected figure known for his work on PyTorch, signals a strategic pivot to stabilize technical leadership. The company must now demonstrate that its vision and research direction remain intact and compelling enough to attract and retain other top-tier talent. Furthermore, it must reassure its formidable list of investors that the departure of founding technical minds does not derail its ambitious roadmap.
Industry analysts often scrutinize such movements for signals about a startup’s internal health and trajectory. Consequently, Murati’s next steps in articulating the company’s strategy and progress will be closely watched. The startup’s ability to innovate and ship compelling AI products will be the ultimate test of its resilience.
Broader Market and Investment Climate
This event occurs within a specific investment climate. In 2025, venture capital scrutiny of AI startups has intensified, with a greater emphasis on path to profitability and tangible technological differentiation. While mega-rounds continue, investors are increasingly vigilant about execution risk. High-profile talent departures can amplify these concerns, potentially affecting future fundraising dynamics for the company and the sector.
The movement also reflects the premium placed on researchers with direct experience in building and scaling large foundation models. As the AI arms race accelerates, this specialized expertise becomes a scarce commodity, commanding significant compensation and influence. Companies like OpenAI, with their operational scale and resources, are uniquely positioned to win back talent they have previously developed.
Conclusion
The departure of two co-founders from Mira Murati’s Thinking Machines Lab to rejoin OpenAI marks a pivotal moment in the ongoing AI talent war. It highlights the powerful draw of established giants and the inherent challenges faced by even the most promising and well-capitalized startups in maintaining a stable, founding leadership team. For Thinking Machines, the path forward hinges on successful leadership transition and unwavering execution. For the broader AI ecosystem, this event serves as a reminder of the intense competition for human capital that will continue to shape the industry’s evolution throughout 2025 and beyond. The ultimate impact on innovation—whether concentrated in a few giants or dispersed across agile startups—remains a key question for the future of artificial intelligence.
FAQs
Q1: Who are the co-founders leaving Thinking Machines Lab for OpenAI?
Barret Zoph, the former CTO and co-founder, and Luke Metz, another co-founder, are returning to OpenAI. A third former OpenAI staffer at Thinking Machines, Sam Schoenholz, is also rejoining.
Q2: Who is replacing Barret Zoph as CTO at Thinking Machines?
Soumith Chintala, a seasoned AI leader known for his contributions to the PyTorch framework, has been appointed as the new Chief Technology Officer of Thinking Machines Lab.
Q3: How significant was Thinking Machines Lab’s funding?
The startup raised a $2 billion seed round in July, led by Andreessen Horowitz, valuing the company at $12 billion. Other investors included Accel, Nvidia, AMD, and Jane Street.
Q4: Is talent movement between OpenAI and startups common?
Yes, there is a well-documented circulation of AI talent between large firms and startups. OpenAI has seen co-founders depart for other ventures, and it frequently re-hires former employees who gained experience elsewhere.
Q5: What does this mean for the competitive AI landscape?
This event underscores the intense competition for top AI researchers and suggests that established companies with vast resources can exert a strong pull, posing a retention challenge for startups. It may lead to increased focus on equity, culture, and clear mission to retain founding teams.