Former OpenAI Chief Technology Officer Mira Murati on Tuesday launched an AI startup, Thinking Machines Lab, with a team of about 30 researchers and engineers from companies including OpenAI, Meta, and Mistral.
The startup aims to develop artificial intelligence systems that integrate human values and expand AI applications beyond those of its competitors.
Murati, who will serve as CEO, is joined by OpenAI co-founder John Schulman as chief scientist and Barret Zoph, OpenAI’s former chief research officer, as CTO.
In a blog post, Thinking Machines Lab stated, “We are building models at the frontier of capabilities in domains like science and programming.” It added, “Ultimately, the most advanced models will unlock the most transformative applications and benefits, such as enabling novel scientific discoveries and engineering breakthroughs.”
The company plans to focus on developing multimodal AI systems that collaborate with users and can adapt to a wide range of applications. In the blog post, the startup noted, “While AI capabilities have advanced dramatically, key gaps remain.” It highlighted the lack of public understanding of AI training processes and the challenges in customizing AI models to specific needs and values.
AI safety will be a core tenet of Thinking Machines Lab’s work.
The company said that it aims to prevent misuse of its models, share best practices for building safe AI, and support external research on AI alignment by releasing code, datasets, and model specifications. “The most important breakthroughs often come from rethinking our objectives, not just optimizing existing metrics,” it stated.
A significant portion of the startup’s team comprises former OpenAI employees, including Zoph, who left OpenAI on the same day as Murati in September. Schulman, who joined Anthropic in August citing a focus on AI alignment, will be the startup’s chief scientist.
The company has been in discussions with venture capital firms to secure funding, according to previous reports. Murati was reportedly in talks to raise over $100 million from investors, though the company has not confirmed this.
Murati joined OpenAI in 2018, leading the development of ChatGPT, DALL-E, and Codex. She briefly served as OpenAI’s interim CEO after Sam Altman’s firing before resigning in October, saying she wanted to “do her own exploration.”
She previously worked at augmented reality firm Leap Motion and Tesla, where she was a senior product manager overseeing the Model X launch.
Thinking Machines Lab is hiring machine learning scientists, engineers, and a research program manager. Murati’s startup is the latest in a series of AI ventures launched by former OpenAI executives, including Ilya Sutskever’s Safe Superintelligence and Anthropic.
