On Tuesday, Mira Murati, the former Chief Technology Officer of OpenAI, announced the launch of her new AI startup, Thinking Machines Lab. The company is backed by a talented team of approximately 30 leading researchers and engineers, many of whom hail from prominent competitors such as OpenAI, Meta, and Mistral.
Thinking Machines Lab aims to carve out a unique niche in the increasingly crowded AI startup landscape by developing artificial intelligence systems that not only encode human values but also target a wider array of applications compared to its rivals. This ambitious vision was detailed in a blog post released by the company on the same day of the announcement.
Murati’s ability to attract top talent from her previous employer is evident, with around two-thirds of the new company’s workforce consisting of former OpenAI employees. Among them is Barret Zoph, a notable researcher who departed from OpenAI on the same day as Murati in late September. Zoph has taken on the role of Chief Technology Officer at Thinking Machines Lab.
Additionally, the startup has secured John Schulman, a co-founder of OpenAI, as its Chief Scientist. Schulman transitioned to Thinking Machines Lab after leaving OpenAI for rival Anthropic in August, citing a desire to concentrate on AI alignment. This concept of AI alignment involves embedding human values into AI models to enhance their safety and reliability, a central focus for Murati’s new venture.
Sources familiar with the situation, who requested anonymity, indicated that more former OpenAI employees are likely to join Thinking Machines Lab in the near future. The startup is also reportedly in discussions to secure venture capital funding from various investors, as previously reported by Reuters.
As the CEO of Thinking Machines Lab, Murati joins a growing trend of former OpenAI executives who have embarked on their own AI startups. Notable examples include Anthropic and Safe Superintelligence, both of which have attracted former OpenAI researchers and successfully raised billions in funding.
Thinking Machines Lab distinguishes itself from competitors through its unique approach, which emphasizes co-design between research and product teams. The company has committed to contributing to the field of AI alignment by sharing code, datasets, and model specifications with the broader research community.
In its mission statement, the startup articulated its goal: “While current systems excel at programming and mathematics, we’re building AI that can adapt to the full spectrum of human expertise and enable a broader spectrum of applications.”
Murati, who joined OpenAI in June 2018, played a pivotal role in the development of ChatGPT and often appeared alongside CEO Sam Altman as a prominent spokesperson for the organization. Her sudden departure was part of a series of high-profile exits from OpenAI, coinciding with significant changes in the company’s governance structure. Before her tenure at OpenAI, Murati gained experience at augmented reality startup Leap Motion and at Tesla.
