Nearly five months after Mira Murati, former CTO of OpenAI, first announced she would be leaving OpenAI after being with the company for over six years, her new venture has emerged from stealth. Thinking Machines Lab, led by Murati as CEO, aims to make artificial intelligence more customizable and accessible to users while pushing the boundaries of AI capabilities in science and programming.

The company has assembled an impressive team of AI veterans, including OpenAI co-founder John Schulman as Thinking Machines Lab's Chief Scientist and former OpenAI chief research officer Barret Zoph as CTO. Schulman left OpenAI shortly before Murati to join the rival AI startup Anthropic and then announced his departure after just five months with the company. At the time of his departure, Schulman did not reveal whether he already had any upcoming plans. Thinking Machines' broader team includes engineers and researchers from leading AI companies who have contributed to prominent projects, including chatbots like ChatGPT and Character.ai, openly available models such as Mistral's, and open-source projects like PyTorch.

Thinking Machines Lab's mission centers on three key objectives:

  • helping people adapt AI systems to their specific needs, by multimodal systems with capabilities beyond coding and mathematics that can support a broader range of use cases and work collaboratively with people (as opposed to chasing autonomous behavior).
  • developing more capable AI foundations with high-quality multimodal foundation models and their corresponding infrastructure,
  • and fostering collaboration within the AI community.

As Murati put it in an X post, "[Thinking Machines Lab's] goal is simple, advance AI by making it broadly useful and understandable through solid foundations, open science, and practical applications." The startup has also committed to contributing to AI safety research by preventing model misuse, sharing industry best practices, and supporting external research through code, datasets, and model specifications releases. In addition to the news of its official launch, Thinking Machines Lab also publicized it is hiring machine learning scientists, engineers, and product builders to expand its team. Details about funding remain unconfirmed, although earlier reports suggested Murati was in talks to raise over $100 million from venture capital firms.