Mira Murati, the former OpenAI chief technology officer (CTO), has unveiled Thinking Machines Lab, an AI startup that aims to develop AI systems that encode human values.
The startup includes a team of around 30 researchers and engineers from companies such as OpenAI, Meta, and Mistral.
Go deeper with GlobalData
The company, which positions itself in the competitive AI startup landscape, seeks to differentiate itself by focusing on AI alignment.
This involves encoding human values into AI models to enhance safety and reliability.
Mira Murati, now CEO of Thinking Machines Lab, in a post on social media X, formerly Twitter, said: “Our goal is simple, advance AI by making it broadly useful and understandable through solid foundations, open science, and practical applications.”
Approximately two-thirds of the team comprises former OpenAI employees, including Barret Zoph, who will serve as the startup’s technology chief, reported Reuters.
OpenAI co-founder John Schulman, who left for Anthropic, an AI safety and research company, in August, is the startup’s chief scientist. Schulman highlighted a desire to “focus on AI alignment” as a reason for his departure.
The startup is reportedly in discussions to raise venture capital funding. It aims to distinguish itself from competitors through a co-design approach by its research and product teams.
This strategy includes sharing code, datasets, and model specifications to contribute to AI alignment research.
Thinking Machines Lab aims to expand AI applications beyond capabilities in programming and mathematics.
“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,” the startup stated.
Murati’s resignation from OpenAI was part of a series of high-profile exits amid governance changes.
