A new startup focusing on AI alignment has been launched by former OpenAI CTO Mira Murati
Mira Murati, formerly the chief technology officer at OpenAI, has launched her own AI startup called Thinking Machines Lab. In a blog post on Tuesday, February 18, the newly established AI research and product company outlined its mission to make AI systems more accessible, customizable, and broadly capable. The startup emphasized that the knowledge of AI training remains concentrated within leading research institutions, limiting public understanding and effective utilization of AI. By addressing this gap, the company aims to develop AI models embedded with human values to enhance their reliability and safety.
Thinking Machines Lab is taking a distinctive approach to AI development, integrating research and product teams in a co-design process. While existing AI models demonstrate proficiency in areas such as programming and mathematics, the startup seeks to create AI systems that can adapt across a wide range of human expertise, unlocking new possibilities for application.
Currently, the company has recruited around 30 AI researchers and engineers, with a significant portion having previously worked at OpenAI. The rest of the team includes professionals from major AI firms such as Meta, Google DeepMind, CharacterAI, and Mistral. The startup’s chief technology officer, Barret Zoph, departed OpenAI on the same day as Murati. Chief scientist John Schulman, who co-founded OpenAI, has also joined the venture. Other key figures include Jonathan Lachman, former OpenAI head of special projects, and Lilian Weng, the company’s former vice president, along with research scientists Luke Metz, Sam Shleifer, and Stephen Roller.
Murati’s career includes previous roles at Leap Motion, an augmented reality startup, and Tesla, the company owned by Elon Musk. She joined OpenAI in June 2018 and played a pivotal role in the development of ChatGPT, frequently appearing alongside CEO Sam Altman. However, in September 2024, she stepped down from her position, citing a desire to create space for her own exploration. Her sudden departure was part of a wave of high-profile exits from OpenAI amid structural changes in its governance.
