
Soumith Chintala Named CTO of Thinking Machines
PyTorch co-creator Soumith Chintala becomes CTO of Thinking Machines Lab, as Mira Murati rebuilds leadership and sets a fresh AI roadmap.
TL;DR
Hyderabad-born PyTorch co‑creator Soumith Chintala is now CTO of Mira Murati’s Thinking Machines Lab, succeeding Barret Zoph. After joining in Nov 2025, he’ll rebuild the technical team and drive the research roadmap as leaders reportedly return to OpenAI—one to track for the ai world organisation audience.
Soumith Chintala’s elevation to CTO signals a reset moment for Thinking Machines Lab as it rebuilds leadership, technical depth, and a clear research direction amid high-profile exits. Below is a detailed, WordPress-SEO-ready paraphrase (with headings), plus meta tags and keyword sets tailored for the ai world organisation.
CTO change at Thinking Machines Lab
Soumith Chintala, a Hyderabad-born AI researcher widely known as a co-creator of PyTorch, has been appointed chief technology officer (CTO) at Thinking Machines Lab, the AI research-and-products company cofounded by former OpenAI CTO Mira Murati. In this role, Chintala is expected to rebuild the company’s technical bench and guide its research roadmap at a time when the organisation is going through rapid internal change.
Chintala takes over the CTO post from Barret Zoph, who has stepped down. The appointment comes after Chintala joined Thinking Machines Lab in November 2025, and it follows the departure of multiple key leaders—including cofouders Barret Zoph and Luke Metz, plus researcher Sam Schoenholz—who have reportedly returned to OpenAI.
Murati announced the move publicly on X, describing Chintala as a seasoned leader with more than a decade of meaningful AI contributions. She also indicated that his focus will be twofold: strengthening the engineering/research team and pushing forward the lab’s broader AI research agenda—an especially high-stakes mandate given the leadership churn.
For the broader AI ecosystem—tracked closely through the ai world organisation, the ai world summit, ai world summit 2025 / 2026 programming discussions, ai world organisation events, and ai conferences by ai world—this kind of executive shift is often a leading indicator of how quickly a new lab can stabilise and start shipping credible research and products.
Mira Murati’s new venture and its momentum
Mira Murati previously served as a cofounder and CTO at OpenAI and also briefly held the interim CEO role during the period when Sam Altman was ousted. She left OpenAI in September 2024 and later started Thinking Machines Lab, aiming to create a mission-driven AI company rather than a purely commercial one.
Thinking Machines officially launched in February 2025 as a public benefit corporation, a structure often used to balance growth objectives with stated public-interest goals. The venture’s profile rose sharply in July 2025, when it reportedly raised a $2 billion seed round at an estimated valuation of around $12 billion—one of the most eye-catching early-stage financings in the AI space.
That kind of capital infusion typically creates expectations for both ambitious research output and credible product direction. With leadership shifting quickly soon after launch, Chintala’s CTO role becomes central to converting that momentum into sustained execution.
Chintala’s background: from open source to Big Tech
Chintala grew up in Hyderabad and studied at Vellore Institute of Technology before moving to the United States, where he enrolled at New York University (NYU). Over time, he became widely associated with PyTorch—an open-source machine learning framework used across academia and industry—which helped shape modern deep learning workflows at scale.
Before joining Thinking Machines Lab, he served as a vice president at Meta, working within the company’s broader AI efforts and alongside its research community. Earlier in his career—between 2011 and 2012—he worked on building one of the fastest AI inference engines for mobile devices, highlighting an early focus on efficient deployment, not just research prototypes.
His open-source work on Torch7 reportedly drew the attention of Yann LeCun, who recruited him to Facebook’s FAIR lab. At FAIR, Chintala played a major role in the development and launch of PyTorch in 2017, after which it rapidly became one of the default frameworks for deep learning research and production experimentation.
Leadership churn, talent pull, and his personal philosophy
Thinking Machines Lab has seen a significant internal reshuffle less than a year after its formal launch. The exits of Zoph, Metz, and Schoenholz—and their return to OpenAI—underscore how intense the talent tug-of-war has become among top-tier AI labs, especially for leadership-level researchers and builders.
This is why Chintala’s new responsibilities go beyond “CTO in title.” Rebuilding a technical organisation during churn requires hiring leverage, cultural clarity, and a crisp research roadmap that can keep existing teams aligned while new talent is onboarded.
Chintala’s public writing also points to a strong belief in open-source and open research—something he connects to growing up without easy access to answers and later valuing the sharing of tools and knowledge as a way to level the playing field. His personal journey includes setbacks and persistence: he has shared that despite a 1420 GRE score, he was rejected by 12 US universities, entered the US on a J‑1 visa before enrolling at NYU, and faced repeated job and visa hurdles before landing at Amazon as a test engineer.


